III.  THE PRACTICE OF FLYTTRÅLFISKE

You can distinguish them at sea, even at great distances.  They appear on the horizon as two little specks which never change relative position.  At night it is the same.  The pinpricks of light from their swaying mastheads are as twin stars low to the horizon, traveling their determined courses, but always the same distance apart.  By these signs you will know they are pair trawlers, fishing together.  But merely to say "together" is not enough; their true function must be expressed in stronger terms.  What they are is two boats acting as one, firmly bonded by the the towing of a common net.  How well they fare in this difficult union depends on the closest cooperation, on complete coordination of every move (Warner, 1977).[1]

Introduction

As the quote above suggests, pair-trawling entails a special kind of activity—one that is physically and socially (and thus also cognitively) difficult to organize.  It is commonly believed (by fishermen from English speaking countries at least, cf. Thomson, 1978) that this organization—and thus effective fishing with this technique—is highly problematic given the stereotype of independent-minded captains who are incapable of long-term, intimate cooperation.  In my experience fishing with a pair from Vindö this kind of cooperation was both "natural" and "effective," although these evaluations are very much complicated by the details of the practice and the changing ecological, economic, and political environments in which it takes place.

Flyttrålfiske[2] (known in English as "mid-water pair-trawling") is a pelagic fishing technique whereby two boats draw a trawl—a sock-like net, suspended between them—through the water.  After several hours (or when the trawl is judged to be full, or departure for port is required) the trawl is hauled on board, the fish emptied into the hull for sorting, the trawl reset in the water and the cycle repeated or departure for port initiated.  One claim often heard among pair fishermen of the Göteborg region is that two boats working together (or better yet, working together with other pairs of boats) can have an advantage over boats working alone.  While this claim is (under certain conditions) true, it ignores many of the complexities involved in this cultural activity.  Modern flyttrålfiske entails the use of two ships, each generally around 30m in length with motor size on the order of 1000hp.  To man each of these boats, while pair-trawling for herring, requires a crew of about 5 men although crew size is generally 6 or 7 to accommodate time off, sickness, etc.  The single most impressive aspect of the flyttrål, both physically and sociologically, is it's flexibility—it's capacity to be used in conjunction with on-line sonar images of where in the water fish and obstacles are located—enabling fishing in nearly all types of water, wherever there is herring.  The practice of modern mid-water pair-trawling along this coast—that is, the on-the-water actions and their historical, social, and cognitive entailments—is precisely what this and the following chapters are about.

Data and analysis

The data for this and the chapter following are derived from a total of seven weeks on board a pair of trawlers from Vindö.  Additional data is drawn from my participation with these fishermen as members of Vindö society, from interviews with these and other fishermen and island residents, from written materials made available by the fishermen's regional and national organizations (Svenska Västkustfiskarnas Centralförbund and Svenska Fiskares Riksförbund) and Göteborgs Universitet, and from my participation on board other fishing boats during my year-long residence on Vindö.  The subject of the next chapter is a "micro-analysis" of fishing activity, constructed from audio and video data recorded while pair-trawling for herring in the Kattegatt straits between Denmark and Sweden.  The purpose of the present chapter is to build the ethnographic background necessary for that analysis and to investigate the cognitive, social, and cultural-historical structural properties of the practice of flyttrålfiske.

A micro-analysis begins with a set of human actions—performances in a practice bounded in time and physical and cultural space—and works toward a set of regularities or principles which relate the actions to each other and the contexts in which they take place.[3]  As with all human actions, the observed set is embedded in a cultural and physical context—actions are constrained by the situation and draw upon resources available within this context for their performance.  Remove or alter the context or (imagine a different) historical trajectory of the practice, and the actions produced take on significantly different forms and meanings.  This is true not simply because different internal structures of the cognitive actors involved are evoked or put into motion by different stimuli.  Rather, human cognition is context-sensitive because human actions are inherently mediated by the mutual organization of internal and external structure and changes in this arrangement introduce fundamentally altered patterns of behavior and meaning.

The difference is an important one.  For the former view (call it the "competence view," in which the environment provides stimuli which simply evoke or activate internal cognitive processes), actions are generated by a mental grammar and are relatively context free.  That is, the explanations given to cognitive events are founded upon the structural/functional properties attributed to be universal aspects of the brain, or "cognitive processes."  Even if it is granted that different cultures provide for different mental contents (thoughts, perceptions, cognitions), culture is generally taken to be simply data which fill the slots operated upon by universal cognitive processes.  The object of cultural analysis, according to this view, is simply to generate a larger sample of possible mental contents in order to understand more about the universal properties of the brain which operate upon them.

For the latter view (call it the "performance view," in which all cognitive acts are mediated by the situation in its historical, material, and social entailments), actions are problematically generated by bringing internal and external structures into coordination—the environment is, in an important sense, half of the material from which action is produced.  Internal structures, according to the performance view, are thus not optimally described as discursively organized knowledge, understandings, plans, propositions, or goal structures whose parsing, search, or deep-structure-to-surface-structure generation is the essence of cognition.  Internal structures are seen, instead, to be learned arrangements of more general information processing resources for performing appropriately in given contexts.[4]  Internal structure, therefore, must be understood with recourse to the patterns of actual performances they enable, in the actual situations where they are carried out, and with reference to the local sets of meanings attributed or evoked in the process.

In summary, a micro-analysis aims to describe human action by paying close attention to the details of performance, as carried out in contexts which recur with enough regularity (as is the case in any socially vital, physically bounded, historically significant practice) to enable the observation of patterned actions.  The goal of such an analysis is to work from these patterns, via exemplary instances, to an account of the constraints and resources (both external and internal, cultural and natural) which structure the activity and the ways in which they do so.  However, before the micro-analysis can be introduced (the subject of the next chapter), there is much background material to cover, and relevant observations to make, about the cultural context and content of the practice of flyttrålfiske.

History of pair-trawling

Trawling for bottom-dwelling fish with pairs of boats is probably as old as the practice of trawling itself.  Industrial pair-trawling for bottom-dwelling species such as hake and cod saw it's debut in earnest in the 1930's in the North Atlantic and the Baltic Sea (Thomson, 1978).  One of the primary constraints on trawling technologies has always been the need for a suitable mechanism for keeping the trawl trimmed as it is towed through the water—that is, a means for providing maximum catching capacity with a minimum of drag.  Using pairs of boats for drawing the trawl is ideal for meeting these constraints.  Each boat provides the needed lateral force to keep the trawl open, and the spatially separated ships afford more efficient use of engine power than a single large ship is capable of providing.

Mid-water trawls have, as the name suggests, the ability to catch fish which spend most of their time swimming or drifting (pelagic fish, like herring) rather than sitting on or drifting near the bottom (like cod, hake, plaice, and flounder).  The invention of mid-water pair-trawling is credited to Yngves Berntsson of Fotö, an island in the Göteborg region, who first experimented with this type of fishing in 1943-44 (Thomson, 1978:77).  The design of the trawl was elaborated into a form—largely unchanged today—by a Dane, Robert Larsen of Skagen (a major Danish port which lies at the northernmost tip of Denmark, 60 nautical miles directly west of Göteborg) in the years 1945-48.  The "mother" of this invention was clearly the long-standing importance of the herring fishery in this region, and the diminishing returns of traditional techniques.

As described in the previous chapter, herring holds a place of special importance in Swedish (indeed, Northern European) culture, making it the catch of choice for hundreds of years in this region of the world.  In the first decades of this century, the technique of purse seining was the dominant means for catching herring.  Imported from the United States via England and Norway, purse seining afforded a natural transition for the vadlag ("net teams") of the west coast of Sweden—from their land-based practice to a mobile fleet of vadfiske ("net fishing") boats—in the last decades of the 19th century.  In the 1920's, as the last sill period receded into history—and the lucrative times of herring fishing with it—fishermen turned to more mobile and less labor intensive forms of fishing.  What began, with the development of the motor and sheltered boat as a way to move a land-based fishing technique (encircle and capture) and its social entailments (large teams of fishermen) out onto the water, soon gave rise to new techniques which exploited the mobility of one's gear as leverage upon the problem of increasing yields while reducing costs.  This is a trend which was to continue into modern times.[5]

The physics and cultural construction of flyttrålfiske

The quotation which opened this chapter described a scene of two boats, fixed with respect to each other on the surface of the water.  In fact, seen at a distance the two boats appear fixed in space, seldom displaying the relative change in position that would reveal motion, as with most boats seen on the horizon.  The boats of a pair move together for long periods of time on precisely parallel courses, apparently attached to each other.  Were we to move closer, the line which seems to connect the boats would twist and turn (now in two dimensional space) as the pair maneuvers around wrecks on the ocean floor and changes in bottom contours or fish population, and act to avoid other boats and pairs in the vicinity.

Figure 15 shows a schematic drawing of flyttrålfiske.  As may be evident from the figure, the "line" which connects the pair of boats is primarily a product of the two captains coordinating their steering so as to maintain a constant distance between their boats.[6]  This is an important aspect of the technique because it fixes one of the variables responsible for controlling the depth of the trawl (see below).  In fact, the general technique employed can be characterized as an effort to fix as many of these variables as possible (ideally all but one) in order to put the trawl at a depth where it captures the most fish possible and yet accommodates quick changes to that depth should obstacles which endanger the trawl and its contents suddenly appear.

Trawls are expensive to replace, and the ocean floor is covered with objects which can inflict damage that could be costly for the teams, in both repair bills and lost productivity.  All told, a trawl costs on the order of $30,000 to $50,000, not including cable (which runs about $3 to $4 per meter).  The obstacles which abound on the bottom of the ocean range from natural protrusions in the sea floor, to discarded garbage such as cars and World War II ammunitions.  Nearly every year a Swedish trawler is reported to have scooped up an old mustard gas bomb while bottom trawling in these waters.[7]  The unfortunate parties involved must then be decontaminated, which can cost the teams several weeks and thousands of dollars in lost productivity and equipment cleaning and replacement.  Probably the most frequently encountered trawling obstacles, however, are the underwater remains of former fishing boats.  Although the identities of wrecks are seldom known, in the year I was in Sweden several fishing boats sunk and many fishermen reported an experience they had aboard a sinking fishing boat.  (Figure 16 shows the density of known shipwrecks in a small area of water northwest of Göteborg.)

Figure 17 shows the dual winches mounted midships on each vessel which, under normal operations, are responsible for setting and changing the depth of the trawl by lengthening and shortening the two (approximately one-inch diameter) cables which stretch most of the distance from each ship to the trawl mouth.  Each of these winches spools cable (vajer or "wire," also known as "warp" in English) in and out to one corner of the four-corner trawl mouth, and is controlled by a single lever next to the captain's chair (see Figure 18).  Thus, the port vessel's winches control the port side of the trawl, via the upper and lower cables which connect to that side of the trawl mouth.  The cables extend from the midships winches aft through pulleys along both gunwales (starboard and port sides of the bridge column) and leave the ship through pulleys mounted up at bridge height in the center (or from both sides) of the ship's stern.

The anatomy of a trawl

Figure 19 shows the basic design and dimensions of the trawl itself.  As shown, the trawl is some 53 m (wide) x 33 m (high) at the mouth, tapering back (almost 200 m) to its end.  The trawl is constructed by sewing together four panels of graduated mesh size.  The mesh size at the mouth is on the order of 3 meters, while at the end (the kalv, or belly, as it is called) the strands of twisted nylon which make a mesh only spread some 30 mm.  This graduation of mesh size has proven effective at increasing the catch of the trawl while minimizing the power required to pull it.  That is, research has shown that even the large mesh sizes effectively "corral" the herring toward the center of the trawl where they become entrapped by smaller meshes (once overcome by the advancing trawl) while permitting other species of fish (with different behavioral patterns) to escape.  In practice, there are many unwanted species of fish or "by-catch" (bifångst) which are pulled in with the herring.  Ongoing research by the (worldwide) industry is aimed at producing techniques and materials for reducing the levels of unwanted by-catch.

The top of the mouth of the trawl ("headline," in English) is mounted with detachable floats and a sonar device (sonde, "net sounder" in English—see Figure 20) which sends a signal via cable to one of the boats.  By convention (and due to the fact that the net sounder cable is spooled out as the trawl is released into the water) the boat which carries the trawl being set is the one which receives the sonar signal.  This signal is displayed by a device on the bridge giving a representation of the fish going in the trawl, as well as the depth of the footrope of the trawl, the fish above and below the trawl, and the bottom of the sea below the trawl (see Figure 21).[8]  The floats on the headline keep the trawl mouth open and are detachable, allowing easy storage of the trawl on a spindle or drum winch.  This winch is mounted midships on the starboard side of the older style trawlers (those with rounded sterns) and all the way aft on the newer trawlers (those with a squared stern and an aft deck open to the sea which sits up quite high off the water, see Figures 17 and 22).

The language of Corks and Stones

The detachable floats of the headline also make the trawl reconfigurable for different conditions and depth ranges.  Along the bottom of the trawl mouth, a chain of lead weights is woven into the primary rope, which holds the lower part of the mouth open in the water.  This same principle, using floats and weights to maintain a vertical spread of netting in the water (and sometimes to establish its depth), is common to all net catching techniques used now and throughout history here.

This long history of homogeneity across fishing practices seems to have played a role in the stability of the terms kork ("cork") and sten ("stone") to identify the floats and weights of all nets, from simple drift nets to huge trawls and seines.  This language originates in times when corks and stones were the actual materials used for floats and weights, respectively.  I take the modern language usage to be more than just "quaint custom."  Rather, I see it as an example of the systematic use of language which mediates the understandings of a community whose members engage in fairly heterogeneous specifics in their daily work routines, yet share some core elements which are the subjects of language use.  Purse seining is really quite different from pair-trawling, and both are different from the practices of retired fishermen engaged in småfiske ("small fishing"), yet all share enough in these separate activities to promote the persistence of terminology whose literally interpreted referential content is non-existent.[9]

In addition, terms for parts of the trawl which may be ambiguous due to the trawl's vertical symmetry are productively modified with the morphemes kork and sten, effectively discriminating the parts by reference to the upper or lower side of the trawl on which they reside.  For example, sten svep refers to the lower ropes which connects the bottom of the trawl to the two vessels' lower cables, while kork svep refers to the upper ropes (see Figure 20).[10]  Thus, these pieces of equipment take on names given by their location in the upper or lower halves of the trawl, based upon the identification of asymmetric aspects of the trawl construction (weights are only found on the bottom, and floats are only found on the top).  There appears, then, to be both consensus-providing and discrimination-making functions at work in the structure of this language use (Clark, 1987; Hutchins & Hazlehurst, 1993).

This view of the persistence of non-literal referential terms—terms which no longer refer to the literal objects from which the names come, and yet are "productive" in the sense that they can be used to define aspects of new technologies—is significantly different from the prevalent views of metaphor in cognitive science.  An often made claim by cognitive linguists (Lakoff and Johnson, 1980; Lakoff and Kövecses, 1987) has been that metaphor provides resources for making abstract concepts more concrete.  Thus, concepts from one domain are used to structure a poorly understood or more abstract domain.  This is fine as far as it goes, but the emphasis shown in most of these analyses entails a commitment to grounding metaphorical structures in a priori universal concepts of bodily experience.  For example, Containers, Paths, and Body Orientation are taken to be grounding image schemas (the fundamental organizers or templates of "concrete" conceptual domains) because they are taken to be universal components of physical experience.

In contrast, the language of fishing nets in Sweden do not reveal universal experience with devices for flotation and orientation under water, but rather the results of engaging in a historical practice and the constraints of communication in and about that practice.  Thus corks and stones—cultural innovations for getting nets to perform properly in the water—have been used to communicate modern properties related to the orientation of underwater nets in fishing technologies because all nets share this feature even though they differ radically in many other respects.  That these concepts are "productive" (that is, entail systematic properties about underwater net orientation and are not simply fixed referential labels) is seen in the way names for other parts of the trawl (the svep lines, as we have seen) take on the kork and sten modifiers.  The point is, that the basis for this systematicity in language cannot be explained by the structure of some image schema such as "Up/Down" as known by direct bodily experience.  Rather, this systematicity in language appears to be the product of experience in (and social discourse about) the practices themselves.  For example, the talk about how to trim the trawls, how deep to set a net (and how to accomplish this), how to fix trawls/nets/seines so they work better, and communicating about the representations (provided by sonar) which actually show the opening of the trawl under water.[11]

How to make a trawl move

Each side (port and starboard) of the trawl is managed by the boat on the respective side, via two cables that extend back from each boat and attach to the svep which themselves attach to the corners of the trawl mouth.  These attachments are made with safety latches which provide quick and easy connecting and disconnecting of the trawl from the ships' cables.  In this way, the trawl (with all 4 svep lines) can be drawn up onto one boat's trawl drum—we can call this boat the "owner" of the trawl.[12]  When the trawl is set in the water, the two svep lines to be pulled by the "non-owning" boat are handed off, and both sets of svep lines are attached to the respective boats cables via the safety latches.

Massive 500 kg. weights (a collection of large iron chain links) reside at the end of each boat's lower cable—where the cable attaches to the sten svep.  These weights, in conjunction with a differential in upper and lower cable + svep length, provide the key mechanism for efficiently controlling trawl depth.  Figure 23 shows a schematic of the steady-state forces that are at work on the trawl, which enable the trawl to maintain a fixed depth in the water while moving forward in the direction of the center of the imaginary line which connects the boats.  To simplify the diagram, only two dimensions are shown in a port-side view of the trawl mouth.  In particular, the lateral components of the forces provided by the two boats pulling outward (thus maintaining their fixed distance apart on the surface of the water), are not shown in this figure.

From this schematic of the steady-state forces on the trawl as it travels forward through the water, we can get an idea of those factors which could lead to a change in the trawl's depth in the water.  Keeping in mind that the only way to effect (and maintain) a new trawl depth is via changes which construct a new set of steady-state forces in the diagram, we can identify a number of actions (intended by the captains or not) which lead to changes in the trawl's depth.  These include: changing the floatation or weight size (not an option while under tow), changed drag on the trawl (more fish, a change in current, or a change to the hydrodynamic properties of the trawl), changing the distance between the boats, changing the speed of the boats, and changing the cable lengths.  The last three changes in this list are those which are possible intended acts which could be taken by captains during the course of drawing the trawl as a means for moving the trawl to a new depth of water.

Avståndet:  Changing and maintaining the distance between boats

In a three-dimensional version of Figure 23, the lateral separation maintained by the boats would be reflected in the "boat" forces coming out of the paper, reflecting "pull" directed along the cables which actually have lateral (port and starboard direction) components.  If the boats pull apart, (and assuming they do not slow down in doing so) the "boat" force is increased in its vertical and lateral components.[13]  This should cause the trawl to rise in the water.  The trawl will come to another steady-state, but now higher up in the water.  The reverse should be the case if the distance between the boats is reduced.

This does not appear to be an effective mechanism for intentionally changing the depth because these forces put unnecessary strain on the equipment—strain which the equipment is not designed to take.  Furthermore, the accuracy with which the captains can control depth changes using this method would be (intractably) dependent upon the length of the cables.  To take up the "slack" of a trawl that is very deep in the water would require many more fathoms of lateral displacement by the boats than when the trawl is drawn near the surface (assuming the trawl is drawn at the same distance aft of the boats).  In fact, the distance between boats is one of those variables that is conventionally fixed—it is the same for all draws, with a given equipment setup, 180 meters or 0.1 nautical miles (nm) for the pair I was with.  As such, there are cultural devices which represent this distance, are used to reason about it, and make it salient and communicable.

First of all, the desired and actual amount of boat separation is known, polysemously, as avståndet ("the distance").  Although avstånd is a generic quantity noun (like our "distance") in this context it is used and understood as a referential noun—it means either the desired or current separation of the boats of the pair, depending on usage.  In both cases avstånd implies something about the expected effects on the trawl, it is not (in this context) an abstract concept of space.  Thus it is not uncommon to hear the warning hålla avståndet ("hold the distance") when the "owning" captain—the one monitoring the net sounder—has detected an unexplained depth change of the trawl, which is assumed to be due to the non-parallel course of the companion boat.  On other occasions, avståndet represents a value which is the actual separation of the boats.  In this usage, the word participates in the more generic meaning of "an abstract distance between two objects" but, more specifically, connotes the distance between the boats of the pair at a given point in time.

Furthermore, the instrument which presents direct sensory evidence of avstånd is radar.  On this monochrome monitor (green foreground on black background) the viewing captain's boat is generally represented in the middle, with the heading of his boat represented by the vertical line straight up to the top center of the screen.[14]  All objects on the surface of the water, within the distance captured by the gain setting of the radar, appear on the screen as echo blips—solid green ovals with axes of about 1/4 to 1/2 inch—whose positions are accurately represented relative to the forward-heading ("up," on the screen) and center-occupying position of the host (radar-sending) boat (see Figure 24).  In order to facilitate reading this screen in terms of avstånd, captains employ an instrument function which generates a dashed circle centered at the host boat's represented position (again, usually the center of the screen).  This circle can be expanded or contracted, by pushing and holding one of two keyboard buttons, to represent some desired fixed distance—in all directions—from the boat.  In this way, avståndet is manifest in the represented circle (fixed at distance 0.1nm from the host boat) and the position of the companion boat's echo blip on or near this circumference.

By convention, the "non-owning" captain is responsible for maintaining avståndet.  This division of labor and responsibilities relieves the captain busy monitoring trawl depth of one thing to worry about, while giving the "non-owning" captain something to attend to.  However, the job of maintaining avståndet requires mutual effort since changes in course, differential boat characteristics, equipment malfunctions and tricky water currents can complicate the job of maintaining parallel courses. Principally, this coordination effort entails ongoing communications so as to make small corrections as needed.  The ships' autopilots facilitate much of this effort by allowing the entry of a desired heading and automatic changes to steering controls which will attempt to maintain the desired course.  During turns and major changes in course, more elaborate manual interventions (often including raising the trawl well clear of the bottom prior to initiating the turn) are required.

Coordinating boat speed

Reducing boat speed also results in a reduction in the "boat" force which will lower the trawl in the water (see Figure 23).  The trawl will come to a new steady-state, now with reduced boat pull and trawl drag, and travelling lower in the water.  The reverse should be the case if boat speed is increased.  In fact, this technique is not employed as a regular method for raising and lowering the trawl.  The primary reason for this is that the trawls are designed to operate at constant speeds.  The force we have simplistically labelled "drag" is really a complex result of the angle and velocity of the trawl with respect to the water.  These trawls have been designed to accommodate not only fish behavior, but also the hydrodynamic properties of pulling them through a range of currents with a range of loads at fixed speeds.  Altering boat speeds would not be a consistent mechanism for changing trawl depth due to the introduction of unwanted hydrodynamic properties affecting the trawl.

A second reason why boat speed is held constant is that it is difficult to establish equivalent speeds between the two boats of the pair (during drawing of the trawl), to a degree sufficiently accurate to assure proper trimming of the trawl.  The boats of these pairs are operated, maintained, and upgraded independently, so there is no mechanism (other than experimental trial and error) for maintaining equivalence between power outputs and resultant speeds through the water.[15]  If, on the other hand, boat speed is not employed as the mechanism for changing trawl depth then the boats need only establish functional equivalence between two or three power outputs (light, medium, and maximum), and these are maintained via meters which display the liters of fuel per hour being burned.[16]

Thirdly, wear and tear on the boats' engines is minimized by driving them at fixed speeds rather than continually decelerating and accelerating.  Finally, a good amount of boat speed is required to overcome the herring, setting a (rather high) lower speed constraint below which it is not feasible to draw and still catch schools of herring.

Skina:  Adjusting the differential in cable length

In fact, adjusting the differential between upper and lower cable lengths is the preferred way to regularly change the depth of the trawl.  As can be seen in Figure 23, drawing in on the lower cable has the effect of reducing the angle ß which reduces the downward pull on the bottom of the trawl.  The dominant effect of this action is that the trawl raises up in the water.  A side effect is that the mouth of the trawl may close down slightly.  On the other hand, lengthening the lower cable has the effect of increasing the angle ß which increases the downward force on the mouth of the trawl, and lowers the trawl in the water.  It should be clear, then, that the "normal" configuration of the trawl (one which puts it in the position to be quickly raised or lowered, while maximizing the vertical opening of the mouth) entails that the lower cable+svep length be greater than the upper cable+svep length.  This difference in length (manifest in the angle ß of Figure 23) is responsible for the downward force on the mouth of the trawl which can be easily reduced in order to effect a (relatively quick) raising of the trawl in the event it is needed to avoid obstacles on the sea bottom or to more productively fish a school of herring.

In fact, since the svep lengths are fixed (the sten svep are 8m longer than the kork svep in the particular configuration of equipment we were fishing with), the only factor worthy of attention by the fishermen is the difference in upper and lower cable length.  This difference is known as skina (a productively used shortening of skilnad, "difference," in the local dialect of the islands).  Both boats must maintain the same skina, in order for the trawl to be properly trimmed, and this is accomplished through a radio channel giving private and undisturbed communications.  This radio channel is a medium for coordinating changes in the trawl depth via actions taken to spool the cables in and out on the boats' respective winches.  A regular feature of the dialog between captains, then, involves the communication of intentions and desires for action regarding changes to the skina.  Since the lower svep is longer than the upper svep by 8 meters, skina values are generally near zero, reflecting a normalization around the most common cable+svep differential of 8m.  A skina  of "five in" thus means a cable length difference of 5 meters, but also translates to a lower to upper cable+svep length differential of +3m.  This configuration of cable lengths does not afford a well-trimmed trawl but is a safe configuration for the initial "set" of the trawl.  Once the pair is under way, those 5 meters will be let out (if possible) in order to lower the trawl to a depth where it is trimmed and fishing productively.

Skina, then, is a word—a culturally provided resource—which "fits" the physical setup of the trawl and the two boats' routines for manipulating the trawl, mediating the cognitive actions entailed in the joint task of maximizing the productivity of the trawl.  Rather than needing to compute differences between upper and lower cable+svep—computations with large values which redundantly involve an unchanging upper cable length and an unchanging svep differential—a single digit is maintained (the value of skina) and dynamically adjusted by simple arithmetic, generally entailing integers 0 through 5 which coincide with the winch actions which affect changes to the length of the lower cable.  For example, spooling out 1 meter when the skina is currently en inne ("one in") gives a skina of zero (usually called lika, "same"), which translates to even cable lengths and an 8m cable+svep differential.  In this configuration the trawl is certain to be well trimmed and, if the net sonar reveals the trawl to be located in the layers of water with the largest concentrations of fish, the captains can be sure that the trawl is fishing efficiently.

The value of skina is maintained in a number of places.  The most basic place is in the represented differential of the cables themselves.  Each cable is marked at regular intervals—beginning with the end which attaches to the svep lines—with additional metal that is woven into the strands of the cable, leaving a readily visible yet unobtrusive clump (see Figure 25).  By noticing the difference represented by any two neighboring markers, one from each cable, skina can be roughly computed (by simply estimating the spread, in meters, between the two markers).  Again, the actual effect on the trawl is a result of the cable+svep differential, but this is just a translation of the fishermen's representation, skina, into the domain of physics for describing the effects of actions taken upon the trawl.  This latter physics need not be salient to the practitioners themselves, but does give us (as analysts) a way to articulate the constraints upon their activity.

Another, more common, method for computing skina entails the winch meter on the bridge which records and displays changes made to cable lengths by the spooling actions taken on the winches via the winch control levers.  Thus, having noted the value of lika (or explicitly initialized the meter to zero) before releasing the cables when the trawl is set, the captain can read the lengths and their difference from this display throughout the draw.  If the meter was not initialized to zero, then skina is computed by subtracting lika from the meter display.  Since the actual values of lika are local to each boat, they are not communicated via radio.  Skina is the value which must be communicated between boats in order to maintain coordination.  When skina is zero, then the word "lika" (but not the value) is spoken, meaning a difference of zero between cable lengths.  Each captain does his own local computations to derive skina from the value of lika and the meter displays or the observed differential in the cable markings.

Setting and drawing the trawl

While charts are provided by trawl manufacturers to compute (theoretical) ideal relationships between total cable length and cable differential for reaching down to all depths of water, in practice fishermen need only know a very small set of parameters which are adequate for establishing the trawl safely in the vicinity of some desired depth.  They then utilize feedback from the sonar device mounted on the trawl to inform them of the appropriate cable lengths for keeping the trawl well-trimmed, clear of obstacles, and fishing the most productive layer of water.  My records show that the general configuration used to set the trawl entailed a basic 5:1 ratio of upper cable length to desired depth of trawl.  Each time the trawl was set with a skina of "five in," and this difference reduced (that is, lower cable let out) once the tow was under way and the safety of the trawl could be monitored as it was lowered in the water.

That is, once a desired depth of fishing has been determined: one of two types of floatation is selected for surface or deep water fishing, respectively; the appropriate amount of upper cable is selected by multiplying by 5 and the lower cable is kept 5 meters shorter; the cables are then adjusted as needed (as revealed by sonar) once under way.  This simple association between cable lengths and depths constitutes a resource for accomplishing the set of the trawl.  But this resource is only functional given the devices which mediate the task of establishing the trawl depth: namely, the net sounder, the cable winches and their meters, the boat sonar which is used to survey the immediate region of water, and the sea charts (or their modern electronic version, the navigator) which enable prediction and suitable "on-line" responses to the situation.  The micro-analysis of the following chapter will investigate how some of these devices mediate actual performance during the activity of searching for where to set the trawl.

In former times, without net sonar, the angle of the cables leaving the stern of the vessel were used (along with cable length) to estimate the depth of the trawl.[17]  (See Figure 26, for a picture of a device designed to measure this angle.  I have no evidence that the exact device pictured was used here on the west coast of Sweden, but I was told that fishermen did compute the cable angles for the reasons stated and I assume some such device was employed.)  This estimation of trawl depth could then be compared with sea charts and computed boat positions to project and plan for the safety of the equipment.  Before the use of net sonar, informants told me, someone was usually stationed on deck in order to monitor the cables themselves for information about trouble.  Collision with the bottom or other obstacles will cause the cables to vibrate, make noise, and even jar the boat if a serious snag developed.

This same deck hand was also assigned the duty of translating the captain's desires into actions taken on the winches.  The use of hydraulics on these ships only began in the sixties, and even for some time after this (before levers were available for controlling the winches from the bridge) crew members were required to hold watch on deck in order to translate the captains desires for changes to cable lengths into actions taken on the winches themselves.  Prior to the use of hydraulics, winches were belt-driven by a continuously running motor whose engagement and disengagement was accomplished by the crew member on deck.  Informants reported that deck watch was not a very desirable job.  It meant being out on deck, or at the ready to go out on deck, at all times.

These tasks have been eliminated from the crew's current routines.  Controlling a vast majority of the draw (virtually everything except setting and hauling the trawl) is now entirely in the hands of the driving captains.  Sonar informs the captains about the need to raise or lower the trawl, and hydraulics—controlled from their respective bridges—effect the changes in cable lengths necessary to meet this need.  These technological changes are part of decades of modernizations which have eliminated the need for manpower, while making the job of being a fisherman less laborious and safer.  In the final section of this chapter, we return to the possible effects these changes have had on the social organization of the crew.

The primary instruments of fishing

Sonar

Sonar works, like radar and in some sense like eyes, as a device which represents information otherwise not perceived, as a map of the objects found within some spatial field or range.  In the case of sonar, this field lies under water.  Unlike eyes, sonar emits a signal which, when bounced off of objects, returns information to a receiver.  It is the differences between emitted and received signal (as opposed to differences between received signal and experience—encoded by evolution or learning—in the case of the eye) which provide the information for constructing an image, which is displayed on some device such as a plotter or a color monitor.  Unlike radar, sonar works underwater, and provides useful information about the reflective properties of objects—like density or hardness.  Thus the phenomena represented by sonar includes schools of fish, the sea bottom, the sea surface, and the location of the trawl in the water.  Each boat has at least two sonar devices: one mounted on the bottom of the boat which projects straight down (ekolod, or "echo sounder") and one mounted on the headline of the trawl (sonde, or "net sounder," or "netsonde") which projects both up and down.

Before the availability of modern color monitors, all sonar displays were paper plotters.  Like other such devices, the paper would scroll by at some fixed rate in the right to left direction, and a stylus mounted on a vertical moving arm would lay down black marks in proportion to the signal received from the logic of the sonar device.  Nowadays, displays are almost exclusively inexpensive color monitors.  Some of the old paper plotters are still in place, remnants of a past technology, on the bridges of these boats.  However, given the amount of time these boats spend cruising around looking for fish, the cost of paper for the plotters is unbearable.  Nonetheless, there remains a skepticism (especially among the older fishermen) for the quality of representation provided by the new displays.  The many colors and the fine gain adjustments of color monitors often provide too much information; information which makes interpretation very difficult.  In Chapter 5, an investigation into the nature of interpreting and communicating about color sonar displays is undertaken.

Since the ekolod transducer captures reflected signals from below (containing one dimension of information about the sea directly below), and the transducer moves forward through the water with some (boat) speed, the resultant image is a two dimensional picture—a side view of the sea covered by the path of the boat on the surface (see Figure 59).  In general, the ekolod does not know about the speed of the boat, so captains (when not towing the trawl), will slow their boats down to some reference speed in order to more accurately interpret the display.

The net sounder's transducers are mounted on the headline of the trawl and send signals both up to the surface of the water and down past the footrope to the sea bottom.  From the returns of these signals, a composite signal is constructed and sent, via a dedicated cable maintained on its own spool winch, to a display on the bridge (see Figure 20).  This winch is usually found high up aft of the bridge where it can spool line over the stern without getting it tangled in the other trawl gear (see Figure 17).  The boat from which the trawl currently in the water came (the "owning" boat, see above) is the one which receives the net sounder's signal.  This signal is displayed on a monitor, and gives an image of the fish going in the trawl, as well as the depth of the footrope of the trawl, the fish above and below the trawl, and the bottom of the sea below the trawl (see Figure 21).

A third kind of sonar device, called asdic, emits signals, not vertically like a ekolod, but from a rotating transducer (as radar does).  This enables the boat to capture information about objects which lie along an imaginary surface (a cone whose apex is at the emitter, and whose height and apex angle is adjustable) rather than an imaginary line (as with an ekolod).  The display of this information is in the form of a circle—the cone which is the field of sensing is projected onto a two dimensional circle which is color coded for aspects of reflection (density, hardness, etc., as with any sonar device.  See Figure 27)

Asdic is primarily employed in purse seining where a more static picture of a three-dimensional region of water is desirable for deciding where to attempt to encircle a large shoal of fish with the seine.  On the move, asdic also proves useful as a sonar device which can give information about fish activity out to the sides of the boat.  The pair of trawlers which are the focus of the next chapter did not have asdic on board.  This technology, in a form which can be used in deep water, is still very expensive.[18]  These captains were, of course, well-aware of the advantages of being able to "see" fish out to both sides of the boat, and not merely directly underneath the boat.

Radar

Radar, like sonar, is an instrument for making salient a spatial field which is (often) not otherwise visible or accurately perceived.  Radar, unlike sonar, provides captains with information about objects above the surface of the water.  These objects are usually other boats but also include land, light houses which emit visible or radio signals, and navigational buoys and markers.  Some of these navigational markers are equipped with radar reflectors for amplifying radar returns, thus making these objects more salient and less ambiguous than they would otherwise be.  In addition, some of the light houses are equipped to generate an identifying signal when activated by radar signals.  The lighthouse signal is displayed on the sending boat's radar display and can be looked up on a sea chart, a list of light houses, or may just be directly known through experience.

As a novice on the bridge, one cannot help but be awestruck with the way captains utilize radar in place of vision to make judgements about actions to take.  As a novice one is overcome by the perception that there is just not enough information available to successfully navigate when, for instance, on board one of these ships leaving harbor.  From the bridge, even if the weather is clear, it is not possible to see out over the towering bow of the ship.  At night, visibility is further reduced by the reflection of inside lights on the greasy, dusty pains of glass which are sparsely distributed (often) around only the forward half of the bridge.  In fact, there are many resources—one learns over time—for aiding in maneuvering the ship which are just not available to a novice.  Some of these are found in the feedback from levers and joysticks which provide thrust and rudder control.[19]  Other resources are inherent in the representations provided by the radar screen about where the boat is relative to certain objects with known or interpretable properties: how steep does that piece of rock pitch into the water (and therefore how close can I come and be safe)? how fast are those boats moving? in what direction relative to me are those boats moving?

There are two primary types of radar: True Motion, and Relative Motion radar.  With True Motion radar the properties of objects captured by the radar are represented with respect to a fixed coordinate system (for instance, latitude and longitude for position and nautical miles per hour for speed).  With Relative Motion radar, the properties of objects captured by the radar are represented with respect to a coordinate system constructed from the properties of the host boat on which the radar sits.  In the latter case, the "host-centered" coordinate system yields representations which take their meaning (for instance, position and movement) relative to the host boat.

In general, Relative Motion radar is older technology that is implemented with many more analog or mechanical computations.  Reflected radio signals received at certain angles and with changes in wave length (compared to the originating signals) are easily "drawn" on the screen as radials originating at the center which change in intensity proportional to the phase shift of the composite signal.  These lines are drawn for discrete positions of the antenna rotation, which yield the characteristic display "sweeps" of the older (typically circular) radar screens.  True Motion radar requires much more digital logic to translate these received signals into various values which can be represented directly or integrated with data collected from other instruments on the bridge.  In general, True Motion radars have Relative Motion "modes" as options, but the reverse is not true—that is, Relative Motion radars have only one mode.

Both types of radar have gain settings which control the sensitivity of the sending signal, and the scale of the represented display image.  Other aspects of these two types of radar differ significantly, and are discussed separately below.  This discussion sets up the background for understanding the micro-analysis of the next chapter.  The two boats of the pair studied in that analysis had different types of radar on board.  One boat had one True Motion radar unit, the other boat had two Relative Motion radar units—one of these was a low-cost unit dedicated to representing avståndet ("the distance," the separation of boats of a pair during fishing, in order to maintain constant trawl depth), as described above.  The discussion below, then, describes not only some general differences between True and Relative Motion radar, but also the particular differences of the two navigational radar units of the pair-trawlers involved in the micro-analysis of the following chapter.

Relative Motion Radar

Relative Motion radar (henceforth, RM radar) displays the host boat in the center with current boat heading "Up," thus constructing a relative coordinate system for interpreting echoes displayed on the screen.  The boat's heading is represented with a vertical line known as stammen ("the stem")—a radial originating in the center and drawn to the middle top of the display screen (see Figures 40 through 44).  As the radar knows nothing of the actual orientation of the boat, there is no representation of this anywhere on this instrument.  Of course, the true orientation of the boat can be retrieved by the captain from other instruments on the bridge.  Likewise, the absolute orientation of other represented objects on the RM radar display can be calculated by a computation which combines the information from the radar with the boat's true heading.

Also available on the RM radar display unit to aid with this kind of computation, are two kinds of "cursors."  One cursor is a point (represented on the screen by a "+") which can be moved anywhere on the display using a mouse-ball.  A numerical field displays this cursor's position relative to the host boat as "XXX˚, YY.YYnm."  This representation gives the cursor's position in polar coordinates relative to the boat's position with current heading as 0°.  Thus "30°, 0.5nm" designates a position 0.5 nautical miles away along the radial which originates at the boat and lies 30 degrees starboard of the boat.[20]  A second cursor, manipulated with two key's on the radar's keyboard, is a radial which originates at the boat and is represented as a dashed line.  A numerical field gives the angle of this radial (in degrees) relative to "Up," and thus relative to the heading of the host boat.  In general, the point cursor is utilized to locate the relative position of point objects in the represented field, while the radial cursor is utilized to consider changes over time in relative position of a moving echo.  That is, the radial can be employed as a single axis against which changes in relative position of some radar echo can be monitored and evaluated.

Finally, the only form of "memory" in RM radar is a kind of "residual trace" left behind by the screen rewriting process.  That is, the sweep which updates the display with the current state of the receptive field can be set to leave behind a residue of former sweeps instead of clearing them away.  This is a display mode which is selectable from the keyboard.  The effect of this mode is, over time, to leave a "trace" of the changing (relative) positions of objects in the radar's receptive field.  From this trace an object's motion can be computed by (mentally) factoring out the contribution made by the host boat's movement.  For instance, if the host boat were lying still, the trace is immediately interpretable as the object's absolute movement in space.  A trace extending away from the host represents a net change in velocity (between host and foreign boat) in the direction (along the trace) away from the representation of the host boat.  If the host boat were moving in the same direction as the foreign boat (and with the same speed), then the trajectory of the foreign boat will not be seen as a trace at all on the radar display.  Rather, the foreign boat will simply be seen as a solitary echo which maintains it's position on the radar screen.

True Motion Radar

True Motion radar represents the host boat and receptive field on a  coordinate system (generally) given in terms of latitude and longitude.  As such, the display can be (and generally is) overlaid with latitude and longitude lines (see Figures 45 and 46).  It should be clear that computing these absolute position fixes (latitude and longitude) is only possible if the host radar "knows" where it is and how it is oriented in order to construct the display.  This is accommodated via input from the ship's gyro compass and GPS (Global Positioning System) satellite navigation receiver (see below).  Thus, all echoes on a True Motion radar display appear to move independently of each other and relative to an absolute latitude/longitude coordinate system.  This "map" of the radar's receptive field is generally represented, like most maps, with "North Up" on the display screen.  The True Motion radar of interest for this study (henceforth, TM radar) was generally operated in a display mode which (like Relative Motion radar) keeps the host boat centered on the display screen.  Thus, while host boat position and orientation was represented in absolute coordinates, movement by the host boat results in the world "scrolling by" this yielding a consistent range of view through time.

As with Relative Motion radar, TM radar has a mouse controlled cursor for dynamically computing the coordinates of points in the receptive field, as represented by the location of the "+" on the display.  On TM radar, the location of this point is represented in an alpha-numerical field in both relative (angle and distance) and absolute (latitude and longitude) coordinates.  Also represented in or near this field is the ship's own position and heading, as received from the interfaced gyro compass and GPS receiver.  Also available on TM radar are both absolute grid lines (latitude and longitude lines) and relative grid lines (concentric circles marking regular distances from the host boat).  Finally, TM radar has a sophisticated set of memory features (even a disk drive for saving information to miniature floppy disks) which accommodate the saving and retrieving of boat trajectories, including the host boat's own history of movements.

ARPA functions

Beginning in 1979, a set of international specifications known as ARPA (Automatic Radar Plotting Aids) was instituted.  The general function of these design guidelines was to equip large ships with the ability to compute and plot useful information about represented radar objects, in order to mitigate the problem of collisions at sea.  These computations include: distance to object, direction to object, course of object, velocity of object.  What I have labelled "True Motion radar" generally has available some subset of these options, which proves to be useful to fishermen who are (generally) very interested in keeping track of where other boats are and where those boats have been.  Although the "collision avoidance" purpose of this technology applies to fishing boats as well as oil tankers, fishermen use this technology primarily to aid their fishing practice rather than plan for safe passage.  (The latter is generally not a problem for these relatively small and maneuverable ships.)

The True Motion radar of interest to this study, the one in place on one of the pair-trawlers of the micro-analysis, had a semi-manual version of these functions.  Any other boat's activity (within radar range) could be traced by moving the cursor to the echo representing the boat and plotting a fix by depressing a key on the keyboard.  In particular, by first selecting the boat number (an assigned number from 1 to 10), then moving the cursor to the echo blip and pushing a key, a point is produced along with a line connected it to the last fix taken for this boat.  The fix, and computed course and velocity of the boat for this segment, can then be viewed.  This manual version of boat tracking was a limitation of the particular True Motion radar unit on this boat.  The automated version of this, in which some number of boats could be tracked in parallel and traces plotted to the screen at regular intervals, was available for purchase but expensive.  Automated fixing and trajectory-plotting for the host boat was available on this particular unit.

What is important about the available manual plotting functions on this particular TM radar unit is that they served as a memory of the trajectories of the other boats.  Recall that RM radar memory is limited to a screen residual of former traces, which goes away when the screen mode is changed or the host boat moves out of range of the currently displayed region of space.  On the TM radar, plotting of the other boats' trajectories is retained in memory as digital information, can be saved to diskette, does not go away with changes in mode setting (unless explicitly requested), and is replotted whenever the region of space of the trajectory comes into range and is projected onto the radar screen.

Navigator

The navigator is an on board computer whose primary function is to represent the position of the ship on a digital sea chart, but also serves as a private and communal knowledge-base for fishing.  This is accomplished via three pieces of technology.  The first is digital sea charts, the second is radio navigation (GPS—Global Positioning System—satellite navigation, and DECCA land-based radio navigation), the third is computer functions which integrate, represent, and remember aspects of the other two technologies as well as interface with the user (and community of users).

Digital Sea Charts

Sea charts are, of course, an ancient technology.  Sea charts represent objects of the natural and man-made world which have proven useful for navigating or thinking about navigating, such as: high points on land, sea depths, latitude and longitude lines, churches, DECCA (a land-based navigation system) curves, light houses and their properties, shipwreck locations and clearance depths, political boundaries, sea bottom composition, etc.

Digital charts are an attempt to take this accumulated store of relevant information and put it in a form that a computer can represent and operate on.  This is a non-trivial problem.  There is an enormous amount of information on a sea chart, and while one way to implement digital charts would be making simple bit-map images of the paper sea charts (thus providing convenient storage and retrieval of the information), this would not be the best use of computer technology.  While the Swedish government (home of the world's most renowned cartographers) has committed itself to standardizing digital sea charts, the fishermen have been utilizing their own versions for several years now.  These are generally created by the manufacturers of the navigators (or subcontracted companies), modified by local knowledge, and handed around among the fishermen.  The standard information represented on these charts is generally limited to: outlines of land, depth curves at a resolution of 30 meters, and some navigational aids such as light houses, and way markers (see Figure 38).

Global Positioning System (GPS) navigation

GPS is a system of some 25 satellites put into orbit by the U.S. government during the 1980's.  Serving both military and commercial purposes, the satellites provide three dimensional positioning fixes at all times, anywhere within the sphere formed by the orbiting satellites.  In general, a fix within this sphere requires 4 satellites which are guaranteed to be found somewhere above the horizon at any time.[21]  In brief, the U.S. government controls use of the system by allowing one channel for commercial use (accuracy to within 100m) and reserving a private military channel (accuracy to within 10m).[22]  A relatively inexpensive GPS receiver can thus continuously (on the order of once every minute) compute a positional fix in latitude and longitude that is accurate to within 100m.  By remembering the previous fix the GPS receiver can easily compute the host boat's velocity and direction as well.  The navigator takes input from the GPS receiver, and plots a fix on the screen, overlaying the boats position—a blinking icon—on the displayed digital sea chart.  Direction and speed are also retrieved and displayed on the navigator screen (see Figure 38).

DECCA navigation

Most of the fishing boats also have (and, prior to the advent of GPS, had exclusively) a DECCA receiver on board, which computes a fix that is plotted with a separate icon on the navigator.  DECCA is a land based navigational system that has been in use in the North Atlantic since the 1940's, and has completely covered Sweden's far and inland waters since 1984.  DECCA, like most radio navigation systems, is built on the principle of an on board receiver translating the differences between signals (sent from spatially separated senders) into locally meaningful coordinates.  The coordinate system of DECCA is based on imaginary hyperbolas whose foci coincide with pairs of signal senders.  This system is truly odd in it's engineering details, but has played an important role in the practice of fishing on this coast in the past, and therefore deserves a little explanation.

In the DECCA system, coverage of some (bounded) region of the earth's surface is provided by four land-based stations, known collectively as a "chain."  Each chain is labeled and consists of one "Master" station and three "Slave" stations.  The slave stations take their names from the colors used to represent their contributions to generating a position fix on a sea chart which has been prepared with DECCA coordinates "Red," "Green," and "Purple," respectively.  Each DECCA fix is given by the intersection of two hyperbolas, each of which is identified by a 4-part label generated by three different properties: the name of the chain, the color of the slave station involved, and a point along the line connecting slave and master station computed from the difference in signals received by the boat from the two stations (see Figure 28).  The fixing of a single coordinate in this system (for example, ignoring the chain name for now, "Red I 16.30" in Figure 28) is based on the ability to measure the phase difference between two radio signals (one from the master and one from the selected slave of the chain) generated at known multiples of a known base waveform frequency (f), and whose origins are some known distance apart (d).[23]  The curve constructed from all points which satisfy the computed difference in distance from the two senders is, by definition, a hyperbola.  That is, the  DECCA receiver fixes the boat's position to be somewhere along the imaginary hyperbola defined by the computed difference of the distances from each of the two stations.  Two such coordinates, involving 2 different slave stations, fix a boat's position at the intersection of two imaginary hyperbolas on the surface of the water.

The meaning of the coordinate system (how the surface of the water is quantified or divided up into meaningful units) was historically determined by the receiver's logic gates, which must be able to discriminate between the received signals in a consistent manner (see Figure 29).

These constraints dictate that, given f—the frequency of the base waveform for the chain—Master always transmits at 6f, Red at 8f, Green at 9f, and Purple at 5f, and this was the standard established.  Coordinates in the system depend upon these ratios, since fixes entail computations over radio signals which must be unambiguously interpreted and represented by the receiver.  Since the base waveform is used to measure phase differences in all of the signals, it is the basis for the primary conceptual divisions of the coordinate system, called "zones" which are labelled with capital letters (usually "A" through "J").[24]  Furthermore, the subdivisions of zones was established by the gating logic of Figure 29—that is, 24, 18, and 30 (for Red, Green, and Purple senders, respectively).  These have become the meaningful conceptual subdivisions of zones (called "lanes") of the coordinate system, and take labels 0-24 (Red), 30-48 (Green), and 50-80 (Purple).

Remembering that two such computations are required to fix the receiver's position, we can now generate a label for the fix which can do some work for a captain by being translated into a point (or, more precisely, into a small region) on a sea chart upon which DECCA coordinates have been drawn.  Working from superordinate labels downward, the fix is identified by: chain name (usually given in the margins of the sea chart since a single chain covers a large region), slave-1 color, zone-1 letter, lane-1 number, slave-2 color, zone-2 letter, lane-2 number (see Figure 28).[25]

In practice, of course, much of this is transparent to fishermen.  What is not transparent is the coordinate system and dealing with it to accomplish navigational objectives.  The bizarre crisscrossing of diverging red, green, and purple lines on an (otherwise orderly) sea chart cannot help but draw an observer's attention.  The DECCA system, despite it's clumsy coordinate system, was a real boon to fishermen of three decades (1950-1980), and is still very salient to older fishermen.  Prior to this, the primary forms of navigation were land based lines of sight, and radio-transmitting light houses from which rather crude bearings could be tuned in.[26]  DECCA, a system whose use was leased out to boats at great expense in it's early years, was a true technological breakthrough for farwater navigation.

The sea chart provides an immediate translation of the hyperbolic coordinates into a point in space which is directly meaningful to fishermen  (see Figure 30).  Although I have only sparse data on how this system was actually used, there are material remnants of its use which can give us clues.  First of all, many of the old private logs containing shipwreck information were recorded in DECCA coordinates.  These are positions where wrecks are known from personal and collective experience to lie, and are therefore positions to be circumvented when trawling.  These positions were generally recorded directly from the boat's DECCA values when a wreck or other unknown underwater obstacle was encountered.  These positions were then later transferred to sea charts (ones with DECCA coordinates overlaid) in order to represent the danger spot in a usable form.

It is instructive to try and imagine what it would be like to make sense of the hyperbolas constructed to crisscross the sea, in terms of translating DECCA coordinates directly into courses of action.  The curves which DECCA coordinates stand for are oriented in directions and with curvature only predictable through local knowledge of them.  Therefore, the symbols read off of the DECCA receiver have little value in terms of their arithmetic "compositionality," or their participation as constituents in a calculus from which solutions to problems may be deduced—they just do not mean anything by themselves.  However, by plotting them on the appropriate chart a representation of the position becomes immediately available, in a form which is immediately meaningful.  Presumably these charts were employed as one part of a cycle of position fixing conducted throughout the day.  (This activity is no longer performed due to the automatic plotting of position fixes—generally, both GPS and DECCA—by the electronic navigator.)

Two uses of DECCA values as raw coordinates did turn up in my interviews and observations.  In each case, these values were employed as parts of an historically-constructed, tool-mediated, activity.  First of all, since older fixes of positions—whether of desirable or undesirable locations—where recorded in the DECCA coordinate system, they are not easily displaced just because the new technology (GPS navigation) is much more accurate.  This is because the continued usefulness of the original observations depends upon the inaccuracies of position fixing due to local deviations which are peculiar to the DECCA system.  Apparently these deviations are not easily factored out, even by machine translation.  Instead, all fishing boats along this coast maintain dual navigation receivers (GPS and DECCA) and continuously plot two different fixes of ship's position on their electronic navigators.  In addition to providing a back-up radio navigation system, this practice renders intelligible many years of observations that have been recorded in terms of DECCA values.  Despite the fact that the icon representing the GPS fix of ship's position is much more accurate, a captain still must worry when he sees the DECCA icon approaching a marker which stands for a wreck because if the latter was entered as a DECCA value then the two DECCA position fixes may be accurate in relative terms (the ship is quite near the wreck as shown) even if they are both inaccurate in absolute terms (the ship and wreck are not where they are represented to be in the ocean).

The use of DECCA values also turned up in interviews with bottom-trawling fishermen.  Prior to the use of computerized navigators, smaller boats—which fish local waters more intensively for sedentary bottom fish—all maintained hand-drawn charts of the trajectories which were known to cover safe (and productive) local waters.  These private databases were, and continue to be, important components of a boat's success.  As the improving technology of bottom-trawls enabled better means for fishing more difficult sea bottoms, these representations—if reproducible—encoded valuable experience.  This experience often came at the expense of damaged and lost equipment during experimentation in the "uncharted" area.  The knowledge thus gained is seen as an investment which is not readily given away to other teams.

These trajectories, curves laboriously reproduced by hand on charts which get rolled up and kept out of sight, are thus valuable assets to the business.  But they are only as valuable as the accuracy of the means for first creating these representations and then turning them back into actual trajectories on the water.  In the pre-DECCA era, this task was generally accomplished via rough rules of thumb which entailed making use of visible land marks—using lines of sight—to establish a starting position, a stopping position, and a heading during the draw.  This method is still evident in these hand-drawn charts.  One sees many plotted trajectories radiating from the dominant visual landmarks of the area.  It's as if the landmarks themselves have served to constrain the possible trajectories that can be represented, and therefore play a role in the trajectories actually taken during fishing.

Also evident on the modern charts are curves, constrained not by lines of sight, but by the hyperbolic shape of adjusting ship's course in order to maintain a fixed DECCA reading on the boat's receiver.  That is, by keeping fixed one of the two DECCA coordinates, a trajectory—a hyperbolic path across the water—is easily and accurately coded and reproduced.  These trajectories are evidenced in the private charts of these fishermen, and the coordination of boat steering actions with the DECCA receiver's displayed symbols is the "algorithm" for translating hyperbolas into a good chance of catching fish.[27]   These trajectories are still used for fishing even though the technology of radio navigation has changed.  They are the residues of actions taken when the practice was constrained by a different technology and different cognitive and social structures for employing it.  However, with the introduction of GPS, digital sea charts, new trawl designs, new markets for fish and new demands placed by debt financing, fishing on more difficult sea bottoms with smaller margins for error has become possible, profitable, and necessary.

Using the navigator for fishing

The navigator's most obvious function is to provide a representational system for navigating the ship.  With automatic plotting of the received coordinates from a navigation system receiver (such as DECCA or GPS) on a digital sea chart, the captain rarely has need to consult a paper sea chart any longer.  Some of the younger captain's rely on this function of the navigator to such an extent that they expressed a bit of anxiety, even if only in jest, over the prospect of having to use a sea chart to reckon their position if the need should arise.  Paper charts were usually pulled out for entry or departure from an unfamiliar port, as the digital versions provide only a sketch of the local navigational details of most harbors.  Otherwise, the paper charts on most of the boats I was on, if accessible at all, were tattered, out-of-date, and virtually unused.  In fact, the main piece of "furniture" on the bridges of these ships, the chart cabinet, is largely obsolete; a monument to the former importance of paper charts.

The other major function provided by the navigator is memory of fished and scanned waters.  The digital charts which get loaded into the computer off of casette or micro-casette sized tapes, can be overlaid with additional information—symbols for trawl obstacles, bottom conditions, fish sightings, and actual trajectories taken during trawling.  Following selection of the plotting mode, the navigator will plot the boat's position at regular intervals, creating a trace of the actual trajectory taken during the draw.  This function is generally turned on when the trawl is set, and turned off when the trawl is hauled, creating a map of crisscrossing and circular trajectories representing a boat's experience in the area (see Figure 16).

This memory function is facilitated by maintaining separate tapes for different regions of water, as required by the memory limitations of the computer.[28]  For the small bottom-trawlers who fish local waters these tapes also contain, like the hand-drawn charts before them, private knowledge of the boat team.  This knowledge entails private strategies for success learned from experience.[29]  The general strategies—what kind of fish to focus on, how "hard" to drive one's enterprise, etc—are not unknown to others, but the details of the geographical terrain covered to accomplish the strategy for a given boat is not generally public.  As such, tapes which contain this information, for example how close one can safely come to a certain underwater berg ("mountain" or raised rock-bed) are not readily distributed.  I was told that captains have available one set of tapes for public distribution, and a separate set for the boat's own use.  This method meets the demands for sharing information placed by general social pressures, while safeguarding the investment in a private fishing strategy.

Bottom-trawlers, then, can accurately reproduce a known draw through the water by steering the boat so as to maintain the boat icon (the boat position as represented on the navigator) on top of the represented trajectory.  These actions entail a radical transformation of the problem of navigating a draw.  The old problem was something like: figure out where to draw, compute target boat positions to accomplish this draw, set course to achieve those positions, continually compute where the boat is with respect to the target positions and generate steering inputs to minimize these differences.  The new problem is something like: generate steering inputs on the joystick rudder control so as to keep the blinking icon on the desired curve represented on the navigator's screen.  Clearly, the new set of tasks has largely concentrated the cognitive actions entailed to those taken by the captain with his navigator and rudder control.

Pair-trawlers, in contrast to coast-near bottom-trawlers, cover much too much water to build up, or bother with maintaining, elaborate knowledge of any specific region of water.  Pair-trawler tapes tend to be much more in the public domain, at least theoretically.  The limiting factor in sharing tapes in this context has more to do with a sparser distribution of these boats across many communities, than any effort to withhold information for it's own sake.[30]  Perhaps most importantly, the nature of fishing for mobile herring, as opposed to bottom-trawling for sedentary species, makes these maps of detailed territory less useful.  Schools of herring do not occupy the same region of water from hour-to-hour, let alone day-to-day.[31]  The technology and practice of pair-trawling is more suited to responding to information on the spot—in order to locate large schools of migratory fish—than it is to reproducing specific trajectories.

Pair-trawlers, therefore, are primarily engaged in establishing their course of actions dynamically.  They are responding to such information as: where are the other boats headed?  how is the bottom changing?  is the fish thinning out?  is the fish thicker on one side than the other?  what are the boats up ahead and off to the sides reporting?  are there any obstacles in our projected path?  is the trawl trimmed and riding through the most productive part of the water?  Much of the navigator's use in pair-trawling therefore entails supporting or providing resources for aspects of this dynamic decision-making process.  Wrecks and other objects plotted on the navigator are discussed between the captains during the draw, and put constraints on the choices of action to take.  The navigator's movable cursor is an important resource for this activity.

The navigator, like the radar display, is equipped with a cursor—controlled by key, mouse, or light pen (see Figure 38).  The represented location of the cursor, in terms of the current scale of the screen, is displayed in a text field in both absolute (latitude and longitude) and relative (baring and distance) coordinates.  Extensive use is made of the cursor, for locating points (waypoints and obstacles) in a representational field (whose background is the digital sea chart) which can be translated into desired headings to take or to avoid.  During a draw, this function is employed to establish joint points of reference between captains of the two boats of the pair.  When not fishing, for instance on the way back to port after a nights fishing, the headings represented by moving the cursor to some desired waypoint, can be read off and dialed into the automatic pilot.  The auto-pilot will then issue the rudder control commands necessary to maintain this heading.  Progress of the boat toward the waypoint is then easily monitored by a captain or crew member on watch.[32]

Radio

There are three kinds of communication channels which are widely employed by pair-trawlers and other fishing boats: cellular telephone, "private" vhf channels, and "public" vhf channels. 

Cellular telephones are used to contact specific parties at long distance who need not monitor their end of the line (other than "being home" when the phone rings!) for a communication link to be established.  Cellular phones are generally employed for calling family, arranging business transactions (such as a buyer for the fish, or supplies for the boat), and reaching specific boats at a great distance.

Public vhf channels are government- or fishermen's organization-sponsored vehicles for coordinating many users of the ocean.  For instance, in the Göteborg region there is a specific channel employed for traffic control inside the busy shipping lanes.  Of most interest to this study is a channel sponsored by the SVC which all of the boats will be tuned in to as a collective vehicle for coordinating fishing and maneuvering efforts.  This channel is used quite heavily to exchange information by boats in the same general area of water.

Private vhf channels are established by individual boats of a pair to facilitate uninterrupted, private, and noise-free communications between the captains of the pair.  Ostensibly, this communication is required for enabling the coordinated actions which ensure safe maneuvering of ships and trawl.  In practice, the privacy of this channel is also desired as a means for pursuing an economic enterprise—that is, making the business of fishing profitable.  This privacy is achieved (or rather, attempted) by installing devices which make available a large number of discrete frequencies—frequencies which can be switched on a regular basis (several times a year) to make it difficult for other boats to listen in on the pair's conversations.  Each boat is also equipped with a scanner which is programmed to pick up the programmed (known) frequencies of other pairs' private channels, or to search for other (unknown) frequencies which are supporting private communications between fishermen.

There also exist "semi-private" channels, that is vhf frequencies with some limited distribution of listeners known to each other.  These channels appeared to exist amongst fleets of boats organized by residence in a community, such as the fleet of smaller boats on Vindö.  The team of pair-trawlers from Vindö (being the only two boats from the island likely to be in a given region of water occupied by the west coast pair-trawling fleet) did not participate (to my knowledge) in any of these "semi-private" conversations.

The practice of pair-trawling for herring

The seasonal and daily cycle of herring behavior

Herring is a migratory fish.  Yearly migration cycles are known (in their general outlines) to cover hundreds of miles and have proven to be quite irregular in their details (see Figure 31).  Fishermen have known about aspects of these cycles for at least a century, and large spawning grounds—where huge concentrations of herring appear yearly—are part of the folklore of fishermen from many lands which border the North Atlantic and North Sea.  Prior to the 1970's, Swedish fishermen also participated in the pursuit of herring in the North Sea.  However, over-fishing (by all of the nations involved) lead to international regulation of these fisheries and the establishment of off-shore political boundaries have virtually eliminated the North Sea fishing option for Swedes.  Nonetheless, as the recent economic chaos brought on by plummeting cod prices attests, herring remains a fundamental fishery for the Swedish fleet.  In general, as shown in Figure 31, herring from the North Sea area will enter into the Skagerack and Kattegatt straits in fall and early winter, leaving again in late winter and spring.  This leaves the Swedish pair-trawling fleet with a narrow window of opportunity (in a small region of water) for catching herring in modern times.

In general, herring lie near the bottom of the sea, relatively inactive, during the middle of the day.  Occasionally, conditions of calm and cold water will encourage the formation of characteristically large, concentrated, mid-water shoals of herring.  These shoals are ideal targets for daytime fishing, especially with purse seining techniques where the entire shoal may be trapped by the catching gear.  It is less often the case that pair-trawling will be productive under these conditions due to the fact that the trawl will effectively disperse the shoal on the first pass, and other shoals are unlikely to be found in the immediate vicinity.  Toward evening, herring become more active and rise up in the water, apparently drawn by light from the night sky.  This fact has been employed by coastal-water seining teams in their use of artificial lights mounted on smaller "lightboats" to attract sprat during the early winter months.  For pair-trawlers, the increased herring activity means the fish rise up where they are more easily spotted and caught by the mobile catching technique of trawling.  Thus night is the primary time for flyttrålfiske.

The cycle of events which constitute a day and week of herring fishing

The general cycle of a 24-hour day of flyttrålfiske consists of: steaming from port to the fishing grounds so as to arrive between 6:00 and 8:00 pm (dusk), conducting two or three "draws," heading for port before daybreak so as to arrive between 7:00 and 10:00 am (when auctions or buyers take fish), unloading the catch, and finally resupplying the ship and making any equipment repairs necessary in preparation for departure again.  A "draw" is itself composed of three parts: searching for fish to decide where to set the trawl, setting and drawing the trawl, hauling the trawl and sorting the fish.  A week of flyttrålfiske on the west coast consists of: supplying the boat over the weekend, leaving Vindö at 12:00 midnight on Sunday, fishing the first night somewhere nearby, fishing the remainder of the week according to the 24-hour cycle given above, and returning home to Vindö Friday afternoon where the remainder of the day is spent cleaning the boat and performing light maintenance work on ship and engine.

Ever since the official formation in 1930 of the west coast fishermen's organization SVC (Svenska Västkustfiskarnas Centralförbund), members have agreed (although not without contentious debate) to abstain from fishing on Sundays and some other holidays (cf. Åberg, 1990).  This particular issue has survived decades of economic, political, and cultural changes primarily due to three factors which, quite accidentally, have acted to preserve this proscription.  These factors are: (1) the strong historical and continuing influence of fundamentalist Christian religion in this area and among fishermen, (2) an emerging consciousness of the problems of overfishing since the 1960's, (3) a tension brought about by the desires to be both a successful fishermen and lead a normal Swedish life (that is, to pay attention to traditional and emerging forms of lifestyles which define oneself not only as a fisherman, but also as a Swede).[33] 

The proscription against Sunday fishing is evidenced, for instance, in the routines of (the few) Swedish ships still fishing in the Atlantic or North Sea.[34]  These boats will often lie still in the water on Sunday—often to the dismay of crew members who may not view rolling about aimlessly on the ocean as a great way to spend their day off.  The proscription against Sunday fishing does not apply when SVC members fish in the Baltic, along the east coast of Sweden, although some choose to abide by it in this case as well.  Among those teams which choose to fish around the clock on the east coast—something which becomes more and more incumbent as idle ships become more and more costly—crew members will be rotated on and off the ship weekly.  This arrangement will generally give each crew member two weeks on board, followed by one week ashore for a 7-member crew of a boat requiring only 5 on-duty members.

Departing from port for the fishing grounds

A day's fishing begins with departure from port for the fishing grounds.  These two locations are generally between two to five hours apart (at 10 knots steaming speed) and will change throughout the week.  Figure 32 shows the locations we fished and landed along the west coast of Sweden during a week in late January, the fourth night of which is analyzed in the micro-analysis of the following chapter.

Arrival at the fishing grounds takes place around 6:00 pm.  By this time, the night time patterns of herring behavior have been established.  Generally, this means that the herring have risen up in the water, due to the nighttime darkness and a behavioral pattern of increased activity at this time of day.  This makes them easier to catch by flyttrål as the fish are clear of the bottom where trawl obstacles and potential trawl damage abound.  This active phase of herring behavior also tends to disperse the fish a bit more at night, but this is not a problem for the mobile fishing technique of pair-trawling.  The choice of fishing area is usually a product of the previous night's catch and information gathering conducted both on the water and in port.  If the catch was good, the pair is likely to return to the same area.  If the catch was bad, then a shift in location is called for and an informal consensus is reached about which direction to move from the reports of boats that fished in the margins of the region.  Other constraints on the decision of where to move to—given that the night before was not satisfactory—include: the proposed distance to the grounds, personal knowledge about how productive this area is for this time of year, and waters of a suitable depth and bottom to support the fish and trawling technique.  The most influential constraint effecting the decision, however, is generated by where the rest of the fleet is headed for the evening.

Among crew members of a pair, this discussion about where to fish is generally confined to the captains who are the ones most aware of the territory fished on the previous night and of the reports made by other pairs about their successes.  It was not so much the case that other crew members were excluded from the discussion, but rather that where the fishing takes place does not really concern them.  Of course, decisions to change strategy which entail shifting to an entirely new region of water and (possibly) alteration to their own schedule of activities do become interesting to non-captain crew members.  Among captains, the discussion is an attempt to pool any known information which might prove useful.  But even here, the procedure is informal, entailing a minimum of explicit planning.  Occasionally, there is an opportunity to probe members from other crews directly, as they walk by on the pier or come over to chat.  But such opportunities are few.  Particularly rare, is the opportunity to engage one of another pair's captains in such informal, face-to-face conversation.  It was clear that the boundaries of this in-port, face-to-face communication were founded upon land-based social ties.  Thus crews from the same port would often be seen associating together during free times in port.

In practice, upon departure from port many of the boats of the fleet seem to follow a core few who have decided where to go (or at least, in what direction to set course) for the night's fishing.  Interestingly, this same pattern of fleet behavior is reported about older forms of fishing as well.  Teams from Vindö, older informants reported, would often wait to see a few "leaders" leaving from the nearby islands, and then follow them out to the fishing grounds.  In the early days of purse seining (ca. 1900), the fleet—once at the grounds—would lie still while small scouting row boats (känare, "feelers") would be sent out to search the waters with hand held lines to determine where the fish density was greatest.  The team's känare was revered for his skill, and presumably for the power he held for deciding (literally, single-handedly) where a crew of 10 to 20 men should spend their next several hours pulling in the huge seine by hand.  This was accomplished on the decks of open boats which afforded very limited range and protection.  Therefore, the "group fishing" strategy was an obvious one for safety, productivity, efficiency, and general entertainment purposes.  Of course, competition among the boats was equally fierce then, the old timers are quick to point out.

I was never very certain of the extent to which our captains (aboard the pair from Vindö) were influenced by—or able to influence—the others' decisions about where to set course for the night's fishing.  My feeling is that a combination of the above strategies (monitoring the radio, knowing the productivity of the night before, talking informally with other crews in port), together with simply following the other boats out of port when they leave, explains most of the decision-making procedure.  Occasionally I heard the discussion directly; our captain yelling down from the bridge to a group of fishermen walking by on the dock (often members of the "core" of the fleet, from the largest west coast fishing port), "So, where are we going tonight?"  In part, this was making small talk.  But it highlighted a lack of explicit independent planning on the part of our own captains—we were going wherever the pack was going!  I tend to think this kind of discourse was more about establishing a relation of inclusion with the influential members of the fleet than about serious information gathering for the purpose of decision-making.

This informally cooperative arrangement among the fleet was also evidenced in the patterns of behavior entailed in actually leaving port.  There was no warning, no radio dialog, no specifically known time to pull out to begin the journey to the fishing grounds.  But every day, sometime between 3:00 and 5:00, a few boats (the "core") would start up their diesel engines and pull away from the docks.  We would fire up our engines (often needing to wake the captain first), scurry around to untie, and get under way as quickly as possible so as not to be left behind.  There was certainly never any real concern among the (non-captain) crew members for where we where headed—they knew we would be fishing somewhere, and the details of where were somewhat incidental.

Finally, the general attitudes associated with this "follow the pack" behavior were also reflected in interviews and more widespread cultural attitudes.  One captain told me that he thought of himself as a follower.  He was not out to set the pace, only to try and keep up with it and all of the payments and responsibilities which go along with that.  He was clearly convinced that being in the same waters as these other boats was a good way to accomplish this.  The micro-analysis of the next chapter will elaborate further on the cognitive aspects of this theme through a detailed examination of the actual means for decision-making regarding the setting of the trawl.

Searching for fish

The trip toward the fishing grounds often begins with (or in the middle of) a meal.  This was generally the case when the unloading (of the previous night's catch) began late or was delayed so as to postpone the mid-day meal until late afternoon.  An effort is made to get the midday meal—the major one both here on board and home on Vindö—cooked and eaten before the ship is underway and rolling about.  The two driving captains of the pair attend to the echo sounder, radar, and radio, while pursuing a course which maintains contact with the pack.  At least one boat of the pair always follows with the pack in order to be in a position to react to what others are doing and the fish they may find.  Being in the "thick" of the pack activity is also important as a means for building ones understanding of where boats are located on radar, in order to make sense of what one hears when others report about the fish they are seeing.  On the trip in to port in the morning the boats may have recorded observations of shoals of fish by entering waypoints into the electronic navigator at the ship's position when the fish were encountered.  Now, on their way back to the same general area, one of the two ships of the pair may pass by these points—generally not far from the course of the pack—to assess the possibilities of fishing there.  This process of recording observations and checking on them later is a form of pursuing private information.

The search proceeds until either good fish are encountered, or it is getting late into the night and the standards of what is "good" have been significantly lowered such that it is deemed worthwhile to set the trawl.  Ideally, the pair would like to get two draws completed in the night.  Each draw can take from one to four hours each, depending on the concentrations of herring encountered.  This means—allowing for some time lost between draws—that the pair needs to set their trawl for the first draw by 10:00 pm in order to meet the schedule of a pre-dawn departure for port.  Of course, this schedule for the night is meaningless without the fish to justify the effort, cost, and risk of setting and drawing the trawl.  It never happened, while I was on board, that we failed to set the trawl at all during a night's fishing.  Occasionally, we would set so late (and on such meager quantities of fish) that we got only one long draw in during the night.  There was definitely an attitude among the captains (particularly the older ones) that we were better off getting the trawl in the water and catching something rather than steaming around looking for better waters all night.

The search to determine where to set the trawl is the most intense period of activity for the two driving captains.  In contrast, once the draw is under way the decisions to be made are highly constrained ones—should we change course, haul in the trawl, change trawl depth, etc.  But where to set the trawl puts the pair's productivity for the day at stake.  The decision, which is left entirely up to the two driving captains, entails information management and negotiation, as the micro-analysis of Chapter 4 reveals in detail.

The sources of information employed in deciding where to set the trawl include: radar displays of the activity of the other boats, visual sightings of the other boats,[35] radio dialog between and with the other boats, sonar displays of the fish encountered, possibly private information obtained about this region (such as sightings the night before), navigator displays which locate the boat relative to features of the digital chart (such as depth contours which identify landmarks where other pairs may be reported to be setting their trawls).  The negotiation of where to set the trawl takes place between the two driving captains.  In fact, the captain who is carrying the trawl which is to be set has the authority to decide by himself.  In practice, he will seldom make a unilateral decision unless the options are unambiguous—such as when a huge shoal of fish is encountered.  Needless to say, this is an exceptional rather than a regular event.

Setting and drawing the trawl

Occasionally, there are other pairs to negotiate with over where to set or draw one's trawl.  In general, the rule is ‘first come first serve' so a premium is placed on getting the trawl set and the course established by quickly getting under way with the draw.  When the decision is made to set the trawl, the crew is awakened or otherwise notified.  Besides the driving captain, there is ostensibly one other (non-captain) crew member on watch at all times (although he may well be below watching TV or chatting with other crew members).  The crew of the boat whose trawl is to be set now has the task of unwinding the trawl from the trawl winch.  This is accomplished by keeping the boat upstream (and under power) away from the trawl, in order to prevent the propellor from becoming fouled.  Setting the trawl is much more easily accomplished in the newer style stern trawlers, as they can slowly pull forward leaving the trawl behind them in the water.  The older style side trawlers have to set the net out perpendicular on the starboard side.  With the ship's hull now perpendicular to the current (and maneuverability largely restricted to forward and aft across the current), side trawler captains must work to avoid drifting onto the trawl.[36]

While setting the trawl, each member of the crew has a role entailing a set of well-oiled routines.  Someone mans the trawl winch controls, two attend to helping the trawl off of the winch spool and checking that it is not tangled, one fetches the floats and net sounder which get attached to the headline.  When the trawl is completely in the water, the svep lines must be attached to the cable ends—two to the host boat and two are handed off to the other boat of the pair for attachment there.  This is accomplished by attaching a casting line to the two svep lines intended for the other boat (which are slack because the trawl ends have been temporarily fixed to the host boat, ensuring that no trawl drag exists on the svep lines during the handoff).  The casting line is a thin cord with a small ball on the end.  The ball acts as a float in case it ends up in the water, but also as a weight for casting the rope to the other boat.  The job for the thrower is to wrap the ball around something on the receiving boat.  A crewman generally waits with a gaff hook to try and retrieve the line if it falls short or bounces back into the water.

In order to accomplish the handoff, the two ships come within 5 meters of each other.  In rough sea it clearly takes a bit of skill (driving these 30 meter boats) to make the handoff and get lined up ready to draw without fouling the gear which is always close by in the water.  These moments are intense ones for captains, even though they have done this hundreds of times before.  Missed attempts at the handoff generally evoke grunts and curses of disgust from the captain—perhaps even yelling out the window—which is noteworthy (for the ethnographer and natives) as an unusual behavioral style.  Seldom does one see much behavior which pits members of the crew against each other as out-standing performers (or, rarer still, as failing performers) of their respective tasks.

Clearly, captains retain a kind of authority and responsibility which sets them apart—but these differences are nearly everywhere underemphasized.  Even the public acknowledgement of a captain's mood-change when in charge on the bridge is given in terms of impersonal job demands ("he's concentrating on all of the information and can't be glib," the captains themselves will say) rather than in terms of authority required or entailed.  That this is the culturally preferred reading of captain behavior is also evident upon hearing negative judgements (among the general public) regarding captains who do a lot of yelling at crew members.  This is behavior which is judged (by residents of Vindö generally) as unnecessary and wrong.  These same attitudes exist among the crew members themselves, although not openly expressed, and these behaviors are interpreted with a bit more understanding about the contexts which engender such actions.

The draw itself proceeds in a fluid, reactive, and yet highly constrained fashion.  The driving captains adjust their course as the fish (seen on their sonar displays) wax and wane.  If one boat of the pair reports more fish than the other, the pair will adjust their course in the more productive direction.  This means that descriptions of fish seen on sonar displays (verbal reports) must faithfully "mean" the same things to each of the captains.  That is, it is important that members from both crews share the terms used to describe the same fish and bottom conditions, as represented on the respective sonar displays.  Each captain will generally produce descriptions of the sonar display spontaneously as relevant changes occur.  If uneven bottom appears, for instance, the "non-owning" captain (the one whose trawl is not in the water, and therefore who cannot see the net sounder display and is not responsible for issuing commands to raise or lower the trawl) is likely to offer up this piece of evidence that the trawl may be in danger soon.  In this capacity, this captain is acting as a "remote sensor" for the "owning" captain, who is in charge of maintaining the trawl depth.  Chapter 5 reports on an investigation into the sharing of language regarding sonar displays, among fishermen from Vindö.

Radio reports are also important pieces of evidence for making decisions to adjust course.  Boats off to the side of the pair are sometimes consulted, and always listened to, in order to determine if a lateral adjustment to the course would be more productive.  Generally, the pair will find themselves drawing in the middle of a pack, with one pair on their starboard and one pair on their port sides.  Often some of these other pairs will be ahead, which gives our captains information about what lies ahead for them.  The behaviors of (and reports from) pairs ahead are taken as evidence regarding the productivity of proceeding forward with the draw.  Pairs ahead which turned around, for instance, were strongly correlated—I informally observed—with the decision by our captains to turn around at or very near this same point in time.

Generally, a draw proceeds along some fixed sea-bottom depth.  A collection of these depths is known as a kant ("edge"), denoting a sloping sea bottom which is usually the side of an underwater mountain or ridge formation.  These kantar often form the boundaries of shoals of herring due to their "collecting" properties.  Shoals of herring are more likely to travel (and collect) along the same depth, where water temperature and other conditions remain constant, than to traverse across these boundaries.  These kantar have representations in the form of depth curves on sea charts—an underwater slope, as on a contour map, is represented by a series of parallel or similar curves.  Captains, responding to queries about their location, will often use the region of a kant as the primary feature of their response.  This region in two dimensional space (now projected onto the surface of the water) is often interpreted, via the speaker's description and by way of the representations given on the map, to constitute a boat or pair trajectory in a certain area of water.

That is, the information in this kind of dialog promotes—for knowledgeable captains—associations which yield relevant inferences (or additional queries) about where boats and fish are located.  Knowing a specific depth that a pair is drawing at gives information about where fish are located that may well generalize to ones own region of water.  Furthermore, knowing the depth at which a pair is drawing yields information about a pair's location and course, since the pair will likely maintain this depth for the duration of the draw.  Finally, if one is following another pair, knowledge of what depth they are fishing at is a crucial resource for determining how to steer to avoid drawing through the exact same water.  If pairs did not tend to maintain a certain depth during their draw, the potential problem—short of much more elaborate boat-tracking systems—would be really difficult to overcome.  If pairs were unconstrained in two dimensional space (on the surface of the water), how could pairs behind be certain they weren't covering water that has already been swept clean?  But since pair behavior is (generally) constrained by depth of the water, knowing the depths being fished by the pairs ahead greatly simplifies the problem for a pair that is following behind.  The pair following behind will often pay attention to the depth of water they (and others) are in, in order to improve their chances of drawing their trawl through undisturbed waters.

The economics of the catch

In general, pair-trawlers are out to catch large sill ("Atlantic herring").  Herring is graded into three classes, based on average fish size, derived from the number of herring per kilogram.  Prices vary (generally) from 2.00 SK/kg for nollor ("zeros," the largest, most desirable fish) to 1.00 SK/kg for the smaller classes.[37]  By-catch (species such as mackerel, whiting, and sometimes cod and other larger fish) must be removed and discarded or sold separately.  An exception to this procedure is when the pair is trawling for skrap ("scrap") or industri ("industrial") fish.  This is fish (of any species but generally price-graded by fat content, and thus preferably herring) which is boiled down and processed into fishmeal for non-human animal consumption or fertilizer.  Skrap fishing is a politically- and ethically-loaded topic, in spite of its long history in this region.  (See Chapter 2, on the history of trankokerier, "fish oil factories," in this region).

When fishing for skrap, the pairs use a trawl which catches all but tiny minnows.  Both fishermen and the general public feel guilt and dismay over the modern practice of cleaning the seas of fish-life with modern technology for the ambiguously useful end-products of fishmeal.  For their part, fishermen are caught in a bind created by economic circumstances which are, at least in the short term, out of their control.  Reduced market prices for sorted herring of high quality, a trend begun in the late 1970's (see Figure 33), currently make the skrap option attractive for owners of larger ships.  Skrap prices are generally (and stably) on the order of .60/kilo (see footnote).  Since much less manual labor is entailed in skrap fishing, and because effective skrap fishing requires large on board storage space, many of the huge ships which have a hard time keeping a large crew when times are tough, turn to skrap fishing as a way to keep the business afloat and profitable.

In turn, the owners of these ships (often under pressure from gigantic debt burdens) have a lot of political clout and use it to keep the skrap option open by lobbying government.  The government is swayed by this clout and a fear of economic and political turmoil if many of these huge ships go under (taking with them the major form of economic livelihood for many coastal communities), and politicians are not inclined to try and shut down the practice of skrap fishing.  Furthermore, the practice is seen as a positive way to assert domestic control over the fate of the fishing industry in general.[38]  As price regulation of consumable fish becomes more difficult in the world economy, supporting the prices for skrap appears to be the best option for easing the tough (short term) economic situation of these fishermen.

The smaller pairs, like the one I was with from Vindö, generally take on board as much sortable sill (i.e., saleable as one of the three classes of sill) as they can, and have a small holding tank down and forward (in the bow of the hull) for about 25 tons of skrap, in the event their haul contains too much by-catch to justify sorting.  Providing that one is landing at a port with a facility for extracting skrap from this hold, the skrap binge ("scrap bin"), the skrap can be unloaded along with the sorted fish when the pair come into port (see Figure 34).[39]  Otherwise, a special landing at a skrap unloading facility must be made later in the week.  Under these latter circumstances, the fish will slam around in the bow of the ship's hull, turning the whole lot into a viscous liquid—which is less desirable from the standpoint of the receiving factories (not to mention the crew which must endure the stench which quickly sets in under these conditions).[40]

The ideal nights fishing entails first packing the containers (big plastic tubs) and lådor (20 and 40 kilo boxes) with nollor; then anything can be put down into the skrap binge ("scrap bin") to fill the boat to capacity.  It is especially undesirable to haul in the trawl with highly mixed fish on the first draw.  Then the decision must be made to either use up some of the skrap binge early in the night, or spend the effort to manually sort the fish—ensuring both that some sorted sill is caught and that some room remains in the skrap binge, affording more flexibility in future decision-making.  Making the decision to put fish in the skrap binge will be based upon: an estimation of the quantity and quality of sill in the mix; the current state of information about market prices; plans for fishing in the remainder of the week, including what the upcoming weather looks like; the physical and psychological condition of the crew in general; and, any specific agreements worked out for purchase of the fish by agents which arrange sales for the fishermen.

If there is less than 50% sortable sill, or if the sill is of a quality which will not fetch a good price, then the fish is likely to be dumped below to the skrap binge.  This is accomplished by simply opening port holes which lead from the mid-deck holding and sorting area to the lower-deck's bow holding tank.  On the other hand, if the catch is to be sorted and ice-packed this begins at once and generally engages all crew members except the driving captain.  The captain is occupied (together with the companion boat's driving captain) in deciding where the trawl should be set for the next draw or setting the course for return to port if dawn is approaching.

Hauling in the trawl

A decision is made to haul the trawl in when: the trawl is estimated to be full; or, there are no more fish to be seen in the area and hauling will allow the pair to search and set in a different area; or, the timing warrants hauling such that another draw can be undertaken or departure for port begun at once.  Determining how full the trawl is, is a non-trivial problem.  Fishing in really tätt ("tight" or, densely packed) shoals can fill the trawl in less than an hour.  Other times, a six hour draw can yield virtually nothing when the trawl is brought to the surface for unloading.

Although it is not a usual event, the pair I was with during the week discussed below in the micro-analysis burst their trawl open one night because—they concluded—the trawl had too much fish in it.[41]   It is possible that this event was precipitated by a weak net joining (a major seam exists precisely where the net was severed), or a bad current, or even contact with an obstacle of some sort.  The two crews had decided in favor of the interpretation that they had left the trawl in the water too long, had collected a massive quantity of fish and—in conjunction with a faulty net seam—the captains (particularly the "owning" captain) were careless in the execution of a turn, causing the trawl to burst.  This was a rather traumatic event because the fish were sparse and it was not a highly productive week, or month for that matter.  There was no blame levied, although the event was attributed to improper handling of a turn in which steps to reduce speed (and raise the trawl)—in order to reduce unwanted torque on the gear—were deemed to have been incorrectly executed.[42]

Two different kinds of procedures, employing very different kinds of resources and computations, exist for making judgements about how full the trawl is.  The newest technology, in use on only a small number of pairs, is a device which measures the stretch of the trawl's mesh at certain points along the trawl.  The stretch of the mesh at these points is registered on the bridge in the form of a panel of lights (thus the term used for the device lamperi, derived from lampa, "lamp" or light).  This technology is not completely debugged, and those who were using it claimed dissatisfaction with it.  In trying to make sense of the discourse about fish in sight and "in the bag"—to use my own expression—I was struck by the frequency with which lamperi are a topic of particular interest for information-inquirers and a topic generally avoided by information-givers.

If we cast the dialog as an exchange of information, an exchange which calls for reciprocation on the parts of interlocutors in radio discourse, then the lamperi create an interesting wrinkle in the normal mechanisms for performing this exchange.  The normal inquiries, as we will see below, are of the form "Where are you?" "Do you see anything [i.e., fish]?"  These speech acts easily fit the mold for expected exchange since reciprocation is automatic (reverse the direction of questioning and ask the same question) and the information is always open to interpretation (affording flexibility in determining the value of ones contribution).[43]  But questions about lamperi—for instance, "Has your lamperi lit up yet?"—both have no discourse reciprocals (the inquirer is unlikely to have a lamperi so the question can not be returned) and are not really negotiable (the answer is either "yes" or "no").  The respondent can choose to lie, of course, but he is constrained in a number of ways which make it unlikely that he will do so.[44]  I became quite convinced that fishermen's voiced dissatisfactions with the lamperi technology—frequently expressed in these radio dialogs in terms such as "it doesn't seem to be working very well" or "we have it turned off because it gives false readings"—were at least partly a product of the social process of information exchange rather than simply an objective analysis (by the speaker) of the functioning of the device.

The second procedure for making judgements about how full the trawl is, relies on estimates of fish seen on the sonar devices.  It will be recalled that this includes the echo sounders mounted on each ship—looking straight down—as well as the net sounder (whose signal is available to one ship of the pair) mounted on the trawl which senses the fish going into the mouth of the trawl.  My own observations suggest that it is quite difficult to determine the productivity of the draw from sonar displays.  While conditions at either end of the productivity scale are not ambiguous—that is, really big schools of herring and no fish at all are two conditions which are easily recognized—the majority of the circumstances fall somewhere in between.  These circumstances are generally characterized by scattered fish, encountered over a long draw, and where the quality (size and species) is very difficult to determine.  In Chapter 5 I attempt to investigate the distribution and nature of expertise entailed in reading sonar representations in a small sample of fishermen who, although they use diverse fishing techniques, all employ sonar images to evaluate the underwater state of affairs.

Finally, we are ready to consider the actions required to bring in the trawl, and unload the catch.  Having made the decision to haul, the two driving captains, in unison, reduce their engine power down to where the ships' forward motions are barely perceived, and begin winding in the trawl cables on the big mid-ships winches.  When the huge 500 kg. weights surface and are brought on board, slack in the svep lines is achieved by backing slowly toward the trawl.  The svep lines are then handed off to the trawl's host boat, in a fashion identical to (but in the reverse direction of) that described above for setting the trawl.  One boat is now (at least momentarily) idle, and will lie still nearby in wait for word about the catch in order to determine what should be done next.

The catch from the draw is usually divided between the two boats, in order to put both crews to work sorting while the next search, draw, or departure for port is begun.  However, this division of the catch has no economic status as all proceeds from sales and costs incurred are redistributed evenly between the two boats of the pair.  Hauling the fish on board proceeds in two steps.  Step one entails the "owning" ship (the one which carries the trawl just drawn) hauling some portion of the catch on board in two-ton increments.  Step two entails the "non-owning" ship coming along side and "taking" some fish (usually the remainder of the catch) on board in similar two-ton lyft ("lifts," see Figure 35).

A lyft is created by winding the trawl netting far enough up onto the trawl winch (on board) to force the herring into the belly or very end of the trawl which remains in the water.  At this point a line from the foredeck boom which has already been connected to a "noose" (a line running through a set of rings sewn into the trawl a few meters from the trawl's end) is pulled taught and used to draw the belly through the water (alongside the ship) toward the foredeck.  (To accommodate this, the trawl winch is wound out to create some slack in the trawl itself.)  This procedure effectively creates a sack of fish (approximately two metric tons in weight) which is lifted out of the water by the boom operator, and swung over to the foredeck fish hatch for unloading.  While the sack is held in this position another crew member unties the knot holding the end of the trawl closed.[45]  When the end opens the fish spill out (hopefully) through the fish hatch in the upper-deck and down into a mid-deck fish sorting area.  The deck hand then reties the trawl end rope and the empty trawl end is lowered back into the water.  The boom line is then spooled out, so that the procedure can be repeated by winding the trawl back onto the trawl winch, pushing more fish into the end of the trawl.

Lyft are first taken on board the "owning" boat (usually half of the catch, as estimated by sight), and then the exact same procedure is repeated for lyft taken on board the "non-owning" boat.  In the latter case the boom line comes from the "non-owning" boat, and requires that the two boats lie in close proximity and coordinate their respective winch actions. 

Sorting the catch

The catch of the draw is quantified in terms of two variables: the number of lyfts (or 2 tons times the number of lyfts) and the approximate percentage of sortable sill.  If the decision is made to sort the catch, the process begins at once.  On the mid-deck stands a machine which does a good portion of this work, but first certain non-herring species are removed by hand from the catch which has come through the fish hatch of the upper deck.  A conveyer belt transports the remaining herring from the lowest point on the mid-deck up to the top of a sloped rack of spaced rails which is the central component of the sorting machine.  The fish land on the rack such that they line up in the direction of the rails, effectively falling into the gaps created by the spacing between the rails.  The entire rack shakes vigorously, under motor power, causing the fish to slide down the slope dropping into bins as the spacing between the diverging rails increases in the downward direction.  Small fish drop through first, and are caught in one bin, large fish drop through last and are caught in another bin.  When a bin fills up, a door opens which spills the contents of the bin onto a conveyer belt running along near the ceiling of the ice and storage hold on the lower deck.

Below, in the storage hold, lådor are filled first with a shovel of ice, then with herring, and topped with another shovel of ice.  The crew member actually loading the herring has motor control of the conveyer belt speed as well as conveyer belt location, allowing him to move the fish to the empty boxes eliminating the need to manually lift any filled boxes.  Despite these effective manual labor saving techniques, packing the herring during rough seas is not an easy chore.  Empty lådor are stacked tight to the ceiling when brought on board, to prevent them from crashing around during the trip.  Now, as the boxes are pulled from these stacks they become unstable and are the source of chaos as the ship slams around in the waves.  In the back and forth sway and violent up and down slam of the bow, one must learn a different physics for tasks as simple as retrieving a shovel full of crushed ice.  A newcomer to this scene cannot help but notice the subtle (and not-so-subtle) body movements that have been developed to cope with this challenge by those who are experienced in this environment.

The captain, for his part, will often attempt to alleviate the burden of those sorting and packing by proceeding with low engine speed and heading off the swell in order to minimize the jolts down below.  From his perspective, however, once the catch is on board, the game is afoot again!  The captain is concerned with where the next draw will take place and (if the sea is calm) will begin to resolve this issue at once, in consultation with the captain of the companion ship, by beginning the search for a place to set the trawl again.

The social organization of Flyttrålfiske

The boats which make up a pair are each owned—as with all fishing boats on Vindö—by two, or three fishermen who are also the skeppare[46] ("skipper(s)," also "master(s)" in wider usage, what I have been calling "captain(s)").  The structural aspects of being, and reasons for becoming, a captain or a crew member are generally the same on pair-trawlers as on other fishing boats, as described in Chapter 2.  What is worth considering here in more detail, following our discussion of the technological means and ecological context of the practice, is how social organization interacts with these, simultaneously shaping and being shaped by the nature of the on board activity.  In particular, the discussion in this section is about specific units of social and cognitive organization—the two crews of the pair, and the pair-trawler fleet as a whole.

Crew organization

It should be apparent that the technology of this practice effectively concentrates the activity of decision-making (at least for the "real time" operation of fishing) in the hands of the captains driving the boats of the pair.  The driving captain has everything he needs at his finger tips for making decisions and acting upon them.  Given the dynamic nature of the activity—e.g., deciding where to search for fish, set the trawl, and/or make a turn with the trawl—there is little time for consulting anyone other than the captain driving the companion boat.  Furthermore, given the distribution of the cognitive work load across instruments and the two captains, there is little need for additional crew members to participate in the gathering or dissemination of information which informs the activity.  Radar and navigator screens reveal and remember boat locations and possibilities for setting and drawing the trawl.  Hydraulics allow arm-chair actions to control the winches which maintain trawl depth and trim.  Radio communications allow the two captains to coordinate their actions during the draw without any participation by other crew members.

Fishermen I spoke with generally attributed this state of affairs to the "rationalization" of the industry—the effort, in the face of changing markets for their product and emerging technologies at home and abroad, to make fishing more efficient.  Although many lamented the fact that fewer people can make a living from fishing nowadays, there was a very strong local perception that the life of a fisherman is much easier today than ever before.  This perception is manifest in the stories and materials (for example, boats) which contrast modern and former practices and make the conveniences (and images of success) in modern fishing salient.[47]

The crew authority structure one might expect to result from this state of affairs—where the power for making decisions lies almost entirely in the hands of the owning captains—is offset by the fact that there are more owner-captains on pair-trawlers than most other fishing boats.  These ships are larger, and require more crew and more owning crew, than the smaller fishing boats used in bottom-trawling along the coast.  But only one owner/captain is really in charge at any point in time on each ship of the pair.  The other, captain(s) of each boat is(are) engaged in the same tasks as everyone else—sorting fish, cleaning equipment, on deck setting or hauling the trawl, or they are below trying to get some sleep.[48]  This division of labor creates two on board roles for owners: driving captain and regular crew-member.  The fact that there are men with the status "captain" who also take on the role of "crew member" is an equalizing factor in the distribution of authority and rights which normally tends to mark the differences between individuals.

In general, the differences between captain and crew-member are minimized—or at least de-emphasized—away from the bridge.  Tasks of responsibility regarding the fishing operation—such as driving the boom or forklift to unload the boat—are distributed across the crew without long-term job specialization.  Individuals do take on specific tasks in the name of efficiency, but do not generally gain any social authority by being specialists since these roles will rotate quite fluidly—not least of all, because crew composition itself is fluid, with members getting one week off every two to four on.  Furthermore, given the age differential possible[49] between a captain and a regular crew member, there are many crucial tasks which a younger man can (and is often known to) perform more proficiently.  This fact also tends to mitigate the effects of differences in authority.  For instance, many of the younger crew are capable of (and willing to) tinker with the modern technology where older fishermen are not.  This skill is a valuable commodity—essential to success of the operation—and serves to highlight the team's dependence upon these more junior crew members.

This deemphasis of differences between individuals is pervasive in life on Vindö—both in the local culture and (I would claim) in national attitudes generally.  Ingvesson, reporting on a fishing community in this same region during the 1960's, remarked upon a general tendency to de-emphasize authority—both on board and in the community (1978).  Captain's on board, she notes, engage in a kind of communication with crew members that appears to entail joint decision-making but which is in fact a process of confirming the captain's proposals.  She sees the same pattern in community meetings, where these same individuals' (captains') inputs carry inordinate weight in spite of the appearance of strictly democratic decision-making.

While non-owner/owner differences are de-emphasized, all things to do with the boat are categorically the domain of boat owners and therefore serve to mark social and pragmatic differences.  Just as boats are a focus of captain/owners' identities (see Chapter 2), they are boundaries between owners and non-owners.  This is clearly established in the division of responsibilities along the line separating "fishing operations" from "boat maintenance."  Thus, a non-owner is not expected to volunteer time to paint the boat during the 2 to 5 week summer vacation—although he might arrange to be paid to do so.  Likewise, maintenance during the fishing season is generally the responsibility of a younger owner who has often received special training to maintain, for example, the boat's engine.

Furthermore, the financial compensation structure of the operation—while noted in the literature and in the local culture for it's socialistic tenets (cf. Hasslöf, 1949:208-210; Ingvesson, 1978:76-77)—in fact makes the owner/non-owner distinction more economically real than is explicitly acknowledged.  All crew members, regardless of experience and expertise, are paid equal shares of the return from the catch after direct costs like provisions and fuel have been removed.[50]  This practice goes back as long as fishing has been collectively conducted on this coast and clearly functions as a status equalizer—it is an institutionalized deemphasis of the meanings of individual differences in actual job routines, experience, and skill.  However, the sum of crew shares is generally only 50% of the earnings which remain after direct costs—the other 50% går till båten ("goes to the boat").  That is, 50% goes to the collective of owners which comprise only 30%-50% of the crew.  While much of this money is often turned back into the collective investment in the business—and thus is seen to benefit the entire crew[51]—owners are in fact financially well-established within the community.  For instance they have, without exception, the newest and largest houses and cars on the island.  My interpretation of this situation, nicely reflected in the phrase "goes to the boat," was that this compensation system and ways of talking about it constitute a set of cultural resources for reconciling inconsistencies in attitudes which strive to de-emphasize differences yet deal with the realities of a practice constrained by: technology, large investments, kinship commitments and real divisions of labor and authority.

It is interesting to speculate about how changes in the technology of pair-trawling may have affected the social organization of the crew in terms of members rights, obligations, authority and power over the years.  First of all, it is important to note that non-captain crew members probably have more free time on their hands now than ever before.  Once fish are on board there are intense periods of work for these men which is centered upon sorting the fish and getting it ready to land at port.  However, while the draw is underway, non-driving crew members have virtually nothing to do but prepare meals, eat, sleep, read, watch TV or sit and chat.  Even the crew member assigned to bridge watch is often found below relaxing or watching TV.  It seems, that is, that regular crew members have fewer demands placed upon them (fewer obligations) now than in the past.  It also seems that, since these fishermen continue to receive the same economic entitlements (rights) that they always have had (equal shares in the catch), that regular crewmen have more power (defined here as rights-to-obligations ratio) in modern pair-trawling than they have had in the past.

Secondly, although captains would appear to have acquired more authority by virtue of having more control over the fishing operation, this authority can only translate into increased power if it is exercised (in the form of demands which increase these individuals' rights) in interaction with the rest of the crew.  But since the technology of fishing has effectively diminished the points of contact between captains and crew, and since the culture puts a premium on minimizing acts which express individual differences and authority, the increase in captains authority does not translate into a commensurate increase in power.  That is, the modernization of fishing—while devastating with regard to the number of individuals who can participate in the industry—has had the interesting effect of making the average fisherman's life much better, at least in the narrow sense given here.[52]

These speculations about the evolution of social organization based upon the structural properties of the tasks involved and how they have changed through the years, are given support by informant reports.  Several informants (now older captains) insisted that when they were young fishermen, the older fishermen in charge (often male relatives of their father) were quite strict, even tyrannical, authority figures.  This picture is in stark contrast to the relationship between older and younger crew members in evidence now.  The current relationship, rather than characterized by a relation of dominance or subordination, is best characterized as a relationship of "joking."  Much time may be spent entertaining (and being entertained by) the youngster's outlook on the world and the way things work, in casual conversation.  Although the tone of this interaction still entails an asymmetry regarding claims of true knowledge about the world (the youngster's naive outlook is the source of many laughs and much kidding), this asymmetry is offset by attitudes which would appear to be significantly different now than they were in the past.  From the older (more authority-holding) crew member's position, there is a definite awareness that the youth's enthusiasm should be nurtured rather than squandered.  For both crew members, there is an awareness that the youth's creative energies are a product of broader (beyond the sphere of fishing) cultural attitudes which empower him to be bold and to speculate beyond the currently known.  The rapid changes older fishermen have experienced within their own practice render this point particularly salient to them.

Finally, modern work routines simply do not involve much interaction between captain and non-captain crew members—minimizing the opportunities for exchange which make psychologically salient to crew members the asymmetries in their positions.  In the past, informants told me, it was categorically the case that the youngest crew member was cook on board.  This position holds one of the few possibilities, I observed, for truly setting a crew member apart from the rest, and labelling the position as synonymous with ‘novice crew member' could only have accentuated this effect.  Work in the kitchen, while symbolically connoting the domestic sphere of life ashore and thus antithetical to life on board, also had practical implications such as: working alone preparing meals while others are engaged in collective work more tangibly connected to the team's productivity; having the responsibility to serve and clean up after others; having additional responsibilities of cleaning and preparation (such as supplying the boat) which are above and beyond work required of one to be a normal crew member.

In modern times, the job of cook falls much more often upon an individual who enjoys performing this activity, and the position—if not occupied permanently by such an individual—will rotate quite fluidly between crew members.  In part, the ability to enjoy this task in modern times stems from the on board attitude of collective "having it good" on the water, which is part of the conscious objectives of boat owners and crew generally.[53]  By contributing to this objective—that is, by being an important component in the process of creating the state of "having it good"—the status of cook appears to be somewhat "elevated" from what it has been in the past.[54]  Finally, it should be noted that many efforts are made to accommodate the additional burdens of the cook by rotating cleanup schedules, self bussing and cleaning of dishes, and collective participation in the selection of menus and discussions of and responsibility for meals.  As in the rest of the practice, modernization has meant that jobs in the kitchen no longer grant exclusive rights nor obligations to single individuals.  Although vestiges of the old system are still extant—where youngest on board gets a higher proportion of the unwanted or menial jobs—the position of cook is not stigmatized as "undesirable," does not fall regularly to the youngest or least experienced crew member, and is made less "different" by collective efforts to reduce the isolation and asymmetries entailed in performing as cook.  It would seem that this "leveling" of the social hierarchy is an instance of how changes in social organization can be predicated upon developments in technology and attendant attitudes which have altered the nature of the practice of fishing.

Pair organization

One major constraint on a captain's actions during fishing, is that he must coordinate them with the captain of the other boat of the pair.  All boat and trawl maneuvers require that actions be taken by both captains—during the draw and in decision-making about where to set the trawl and where to look for fish.  This coordination is facilitated by the dedicated vhf radio channel—the pair's private channel—as described above.  Each pair has such a private channel, and all boats have radio scanner's to try to listen in to these "private" channels!  The private channel is justified in terms of the need for immediate, uninterrupted communication between the two boats.  However, it is also a device for protecting the economic and social interests of their collective enterprise, and for making the job more interesting.

In general terms, the pair constitutes a single economic unit.  All costs and earnings are shared evenly between the two boats.  Earnings, of course, are a result of the fish caught and the price received for those fish.  Both of these activities then—finding and catching fish and organizing their sale—constitute the primary economic cooperation of the pair.  Communications over the private channel entail much discussion about these activities, and are deemed worth protecting.  Although they travel together with other boats as part of a fleet (see below), if the pair has some information that will cost them earnings if others were to know about it, then they are interested in keeping that information to themselves.

Of course, this paints a picture of pure "economic rationality" that does not, in practice, exist.  Information about a possible state of affairs does not translate into that state of affairs directly.  Fish can disperse from where they were once seen, prices for and commitments to buy fish change at the last minute, "a lot" of fish may be less than what others have found elsewhere, etc.  In reality, there is a much higher degree of uncertainty about maximizing the "earnings - cost" equation, which requires skillful manipulation of the cognitive and social variables which are involved.

Some theorists have utilized "Game Theory" as a way of characterizing this kind of economic situation.  This theory assumes—I think correctly, and certainly the fishermen hold it to be true—that each "player" (economic unit) is concerned with maximizing the "earnings - cost" value of their operation.[55]  Where game theory fails is in the way it casts the actual pragmatics of this activity.  For instance, it is clear that players may choose a strategy which "looses ground" in the short term yet makes long term gains.  Although sophisticated models may attempt to accommodate this twist, the theory generally fails to relativize what "costs" and "earnings" may be to the players themselves.  That is, without mapping out all of the terrain (the means for action in the practice) the analyst is left with a very weakly explanatory, and descriptively misleading model.  The higher-level agent goals which may be reasonably represented by the analyst's "payoff matrix" (the formula for assigning costs and earnings to players choices for action) do not directly shed any light on mechanisms operating at the level of human actions situated in a specific cultural context.[56]  One goal of this dissertation is to make explicit the cultural and cognitive means responsible for the phenomena of cooperation and competition among boats of the pair-trawler fleet.

The history of collective effort by the two boats of a pair is usually quite extensive, although only loosely organized.  The pair I was with had been fishing together for almost twenty years.  However, these boats will regularly (throughout the year as prospects for herring fishing fluctuate) split up to pursue bottom-trawling for cod, individually, with no lingering commitment other than an agreement to agree to begin fishing together again when the time is right.  Furthermore, pair composition is somewhat fluid, as evidenced by arrangements regularly made to team up with other boats when ones companion boat gets sold or is in for long periods of repair.  Thus the pair's contract together, while enduring over many years, is flexible throughout each year.

The pair's mutual dependence during some 50% of the year—as well as their locations of residence in the same or neighboring communities—does generate a degree of cooperation throughout the year that is not extended to all other boats.  Private communications and a free flow of information between the two boats, even while they are fishing individually as bottom-trawlers, will be maintained whenever possible.  It is also usual for the two boats to run errands and perform other time- and cost-saving measures for each other, even during those periods when the boats are operating as separate economic units.

The most intense requirements for coordination between the two boats come during the draw itself.  The reasons and means for this coordination have been discussed above.  The boats must maintain a fixed distance (avstånd) and must take parallel actions on the winches to effect changes in the trawl depth.  To accommodate this process, there is a division of labor which gives the two driving captains asymmetric responsibilities.  The ship that carried the trawl which gets set in the water, can be called the "owner."[57]  The "owning" captain is the one in charge of the draw.  He is the one who receives and views the net sounder signal; he is the one who commands actions for changing the trawl depth; he is the one who has the ultimate authority about decisions to change course or to stop and haul in the trawl.  The other captain's primary responsibility is to "hold the distance," that is, to guarantee the maintenance of avståndet.  In practice (as we will see in the micro-analysis), decisions are not made unilaterally.  The one exception here is trawl-depth changes, for which only the "owning" captain has suitable information—namely a view of whether the trawl is in danger or is fishing the most productive layer of water.

Fleet organization

What I am calling the "fleet" is a loose collection of pairs, between 3 and 10, which are fishing the same region of water.  All of these teams are members of Swedens Fishermen's Organization (SFR, Sveriges Fiskares Riksförbund) and most come from the Göteborg region and are members of the Swedish West Coast Fishermen's Organization (SVC, Svenska Västkustfiskares Centralförbund).  These teams find themselves in the same region of water on an informal, yet reliably predictable, basis.  First of all, their location in the region follows general knowledge about the migratory patterns of the herring.  Second, their membership in a given week is the result of a dynamic process which begins with membership from the previous week: boats change gear and move to other waters to pursue something else if perceived profitability drops (and there is a better alternative); boats leave for repairs; and boats arrive because they have heard or imagine it to be the case that success can be found here, fishing with the fleet.

In addition there are other nations' boats—usually Danish, in the waters between the two countries—which often follow along.  The Swedish pairs all use a public radio channel, whose frequency is owned by SVC, to communicate with each other.  The functional properties of this collectivity (the fleet) are complicated (I believe)—although somewhat transparent to the fishermen themselves.  From their standpoint, this kind of collective (fleet) fishing is a completely rational basis for pursuing their business.  "More boats can catch more fish (per boat) than the boats can catch fishing alone," the fishermen told me.  There is much truth in this claim.  The scope of one's instruments for seeing fish is largely restricted to the one-dimensional trajectory that the boat takes through the water.  Multiple boats increase the chances of coming across schools of fish, each of which is generally larger than one or several boats can take on board in one night's fishing.

But, as with many cultural activities, both the means and the meaning of this practice are not adequately described by this claim about rational behavior.  First of all, while the teams do flock together it was not at all clear that the fish we were catching were a product of collective "effort" on any given night.  The general search strategy was much more one of: (1) hang with the fleet while privately discussing possible options available by drawing upon accumulated information; if none of these options are pursued then, (2) when other boats are making decisions to set their trawls, evaluate the choices of (a) setting here along with them, or (b) continue on because what is seen here is not good enough, which might mean (c) setting somewhere else based on private information.  In the micro-analysis below, we will look at some actual instances of this decision-making process and elaborate on its form and underlying mechanisms.

Secondly, "collective action" may be more aptly characteristic of fishermen behavior on the time scale of weeks and seasons—as channelled through the institutional means of their political organization—than behavior which is explicit during, for instance, the search for fish on any given night.  The fishermen's organizations provide important (and powerful) political clout, both on the national policy front and via collective bargaining with local buyers and their organizations.  However, strategies for improving one's business over the long term also include explicit reciprocation in the form of exchange of information along the way.  The forms of this exchange are importantly shaped by the constraints upon knowing and acting which are inherent in the activity itself, as the following chapter will demonstrate.

Finally, face-to-face (or radio mediated) acts of exchange are not valued by participants solely in economic terms, but also in terms of benefits to individuals received in the form of self-confirmation about who one is.  These benefits to individuals are attained via recognition that the members of these other teams are more like oneself than not.  This recognition is brought home every day by acting the part—including performing according to the conventionalized codes of conduct which are publicly monitored—and reinforced through collective rituals sponsored by the fishermen's organization.  That is, through common experience (brought about by conducting the practice so as to create common experience) fishermen identify themselves by reference to a collectivity, and actions are taken which simultaneously fill the expectations of others and are self-rewarding.  This process yields forms of behavior which are constrained not only by the potential for catching fish but the potential for achieving success as defined by what it means to be a fisherman as constructed through participation in the practice.


 



[1]In reference to the Spanish pair-trawling fleet.

[2]Grammatically, flyttrål is ambiguous as to whether it is derived from the verb flytta (to move) or the verb flyta (to float).  Both verb roots, "float" and "move," can be transitive or intransitive, and both accurately describe the technique.  I was always more influenced by the latter etymology, seeing the action of the trawl as caused by the actions of the fishermen, and this "moving" reflected in the sense of the term.  However, I was told by one captain in later correspondence that the root "float" is most definitely the correct sense for the term.  In retrospect, this meaning is reasonably accounted for by the fact that the practice is defined in contrast with another popular form of trawling, namely bottom trawling, which is the meaning of the unmarked form trålfiske.

[3]It should be noted that by "actions" I mean intentional behaviors—behaviors that are both teleological (driven by expectations about their consequences—they are "intended" behaviors) and sensitive to context.  It should also be clear that a micro-analysis requires much more information than just recordings of behaviors—it requires many cultural materials (interviews, written histories, and documentation, knowledge about this and related areas of activity, etc.) in order to translate physical behaviors into meaningful human actions.

[4]This definition of performance is no more question-begging than the competence view's own definition.  Both rely on some notion of coherent actions—what I have called "performing appropriately"—as the target which the research is meant to explain.

[5]This rather brief exposition should not give the impression that trawling represents a culmination of this process, nor that this particular evolution is a natural course of affairs predicated upon economic or other laws.  Hasslöf (1949) does much more justice to this history than can even be attempted here.  For instance, he reports that while the steam engine afforded fishermen an earlier opportunity to go mobile with their practice, it never came to pass because the use of such ships required financing and ownership from outside the sphere of influence of fishermen themselves and was thus never adopted.  In contrast, the gas motor made possible a technological adaptation that was easily incorporated into the existing social organization of fishermen and fishing practice.

[6]Thomson (1978:103) reports that some smaller boat pair-trawling techniques (he notes French, Belgian, and Dutch teams) do use a bow line to connect the pair, thus mitigating the chances of straining gear—and possibly altering the nature of the steering task for the captains.

[7]Thousands of these bombs, loaded onto vessels, were intentionally sunk by the allies following WWII as a means of disposal.  This activity has only  recently "resurfaced" in the context of multi-lateral efforts to rid the seas of sunken environmental time bombs.

[8]As discussed below, this division of labor brought on by the fact that there is only one sonar signal, generates a certain cognitive and social organization whereby one boat (below called the "trawl owning" or "owning" boat) is "in charge" during the set and draw of the trawl.  This role generally rotates back and forth between the boats following each draw.

[9]The fishermen themselves seemed a little embarrassed by the terms.  Apparently, in their eyes the terms (when consciously examined) do reveal "quaint custom," or at least some were convinced that this would be my perception and it made them uncomfortable, or gave them cause to laugh at themselves.

[10]Svep itself appears to be derived from the role this rope (more precisely, the lower ones or sten svep) play on bottom trawls, viz. they sveper  (literally, "sweep")  the fish up off the bottom and chase them into the center portion of the trawl mouth.

[11]As a final note, I should mention that in English these same parts of the trawl do indeed relate to bodily experience, or at least representations of the body.  Thus, the top of the trawl opening is known as the "headline" and the bottom as the "footrope."  There is some evidence that Swedish fishermen also ascribe properties of a biological sort to the trawl.  While the catching end of the trawl is called the kalv ("calf," or "veal," perhaps connoting this parts role as the "meat producing" end), I was also told it is called the belly (which I believe comes directly from English, "belly") perhaps connoting this end's role as the place where fish end up once they are "eaten" by the trawl.  Unfortunately, I have no additional data for pursuing these speculations.

[12]The terminology of "owner" and "non-owner" is discussed below.

[13]One can imagine the "boat" vectors swinging out toward the reader, and extending (at least momentarily) in length as the boats pull apart.

[14]This discussion assumes the radar is set up in Relative Motion mode, with Heading Up selected.  The differences between True and Relative Motion radar are discussed below.

[15]Such would not be the case, for example, if the pairs where company owned and built as pairs, as is the case in the Spanish fleet (cf. Warner, 1977).

[16]This method also accommodates better financial control of the operation.

[17]This is only possible in good weather, and even then can not account for cable sway, unusual currents, etc. which distort the idealized geometry required by this technique for accurately computing the depth of the trawl.

[18]I was told it would cost between $100,000 and $200,000 to have a suitable asdic installed.

[19]The old wheels of these ships, while still extant, are usually built into a corner behind some add-on piece of technology, or often used as a convenient foot rest.  I was never really sure whether they functioned at all as backup rudder control, although I noticed many of them did move feebly when the ship was being turned by the other control levers or the auto pilot.

[20]All compass directions in the Swedish system are based on a 360 degree circle swept in the clockwise direction.

[21]Four satellites are needed as follows: one for each spatial dimension plus one to correct for timing errors.  Since the fishermen are only interested in two dimensional fixes (latitude and longitude), their receivers are somewhat simplified.

[22]The fishermen enjoyed telling me how their systems were dead accurate during the Persian Gulf War.  I have no explanation for why this should have been the case other than that the US military had turned off the "noise" on the commercial system.  I did notice the general interest in this phenomenon (and the telling of it)  by the fishermen and attributed it to a kind of "vicarious participation" in the technological wizardry that was so (skillfully) promoted during this war.  Near the end of my stay, a ground-based Scandinavian "corrector" was coming on-line which significantly increased the accuracy of the GPS system in these waters (called Differential GPS, or DGPS).

[23]All of this "knowing" is accessible to the receiver once the particular DECCA "chain" (the label for a "Master" and 3 "Slaves") has been dialed in (see Figure 28).

[24]The actual distance of a zone is relative to the wavelength of this base waveform, call it µ (µ = 2 * speed of light / f).  Thus, the number of zones can be computed from the number of times µ can fit between the two stations, or d/µ, and varies (in general) for every Master/Slave pair since d varies within each chain and µ varies between chains. 

[25]Actually, lanes are usually divided into hundredths or "centilanes."

[26]These worked on the principle that a receiver could detect the direction of the sender, by the angle of the receipt of the radio wave from the sender.

[27]It should be briefly mentioned that while boats generally have both a DECCA and a GPS receiver on board, and continually representing the boat's position on the bridge's electronic navigator, the GPS receiver is generally much more accurate.  This fact is readily verified on a continual basis by observation of the discrepancy in the displayed position fixes, together with independent evidence for where the boat is (e.g., tied up to the dock).  One source of the DECCA system's error is readily seen in the hyperbolic coordinate system.  Hyperbola's diverge from each other at distances removed from their common focus.  Therefore, any finite resolution (or area of uncertainty) in the fix near the focus will be larger at distances away from the sending stations—even before the practical aspects of this distance (e.g., increased noise in the received signals) is taken into account.

[28]Memory limitations of the computer determined how large a region can be represented at one time and how much overlaid information can be accumulated, since it all must be loaded into memory from tape for use.

[29]Although this experience is also handed down between generations, it appears to be the case that changes in technology and fish markets quickly make the specifics of this information obsolete.  This was probably not the case in former times when more slowly changing technologies made the intergenerational transmission of experience more important.

[30]There was only one "pair" from Vindö, while there were more than a half dozen competitive bottom-trawlers from the island.  The opportunity (and motivation due to social pressure) to share tapes among pair-trawlers was therefore much lower.

[31]However, from season-to-season fish are known, as reported in a previous section, to return to the same general areas.  Therefore experience in specific areas as represented on these tapes do provide a decision-making resource in the most general sense.

[32]Navigators are also able to take inputs from sonar and to generate outputs (automatically) for an autopilot.  Neither of these interfaces were functional on any of the boats that I spent time with.

[33]This latter constraint is perhaps best illustrated by an oft-heard expression of concern by Scandinavian fishermen generally for the migration north of Spanish and other Southern European fisherman under the new European Community alliance.  This concern expresses a logic whereby the (understood to be) nomadic lifestyles of these other peoples will create an economic imbalance (as far as fishing goes) because Scandinavians—I was told—would not be interested in fishing so far from home so as to exploit Southern European waters.

[34]Through agreements with the EC and other nations which border these waters, Swedes still maintain a few licences for fishing in the North Sea.

[35]When the boats stop to set, they turn on deck lamps which are readily seen at some distance on clear nights.

[36]During my time on board the rear trawler of the pair, we were once called over to tow our companion ship, a side trawler, upstream and clear from her trawl.

[37]These market prices reflect only 1/2 to 2/3 of what the fishermen actually made from the sale during the economic climate of my residence.  The difference was made up by government price-supports, which were the subject of contentious debate in the country (Statens Offentliga Utredningar, 1989).

[38]However, as has been pointed out by the advocates of reform, the price supports for skrap fishing seem to benefit only a small proportion of the entire Swedish fleet (cf. Statens Offentliga Utredningar, 1989).

[39]This piece of equipment is a mobile conveyor belt that has a vertical component which can be lowered down into the hull of the docked ship (see Figure 34).  More importantly, the receiver of skrap must have a transportation or processing facility for dealing with the fish.

[40]This phenomenon is not only unpleasant, but also dangerous.  In the summer of 1993, two cases of death were reported among Danish laborers who unload the ships' skrap binge, due to inhalation of toxic gases which quickly build up under these conditions (Yrkesfiskaren, 1993).

[41]Unfortunately, I was asleep when the event took place and felt, personally, a certain shock or trauma over the incident which did not precipitate the necessary intensity of questioning—on my part—to make complete sense of this rare event.

[42]It was expressed to me by several fishermen that dwelling on past events, mistakes or unfortunate incidents, is a waste of time.  There is a general attitude of optimism entailed here, about looking forward to possibilities rather than back on failures.  This attitude also fits with a denial of levying fault on individuals—which would foreground them as individuals with certain failings and potentials—which is in line with theme which runs throughout the culture here.

[43]An exception to this is the question "What are your values [latitude and longitude readings from the GPS receiver]?"—which must be answered unambiguously.  If the argument I am making here about lamperi introducing a new constraint upon this process, then it may have been the case that the introduction of automated navigation like GPS and DECCA had similar effects in the past.

[44]Here I am thinking of his conscience, social pressure from other crew members who are witness to the public behavior, as well as the chances that his lie will be discovered—either by word of mouth about the behavior, or by deduction once the results of the nights catch are publicly available.  Following hauling of the trawls, something which happens more or less in unison across the fleet, there is nearly always a round of radio exchange regarding the catches made by each pair.  This "accounting," together with the physical evidence produced in port the following day, appeared to make blatant deception an unworkable option.

[45]It is a non-trivial task in rough seas to time it so that this knot is undone when the swinging sack is stably positioned above the open hatch—all the while avoiding being crushed by two tons of swinging herring and maintaining ones footing on the deck.  Furthermore, there is a certain urgency involved since each swing of the lyft, as it crashes into the 3-sided rail surrounding the open hatch—the purpose of which is explicitly to stabilize this pendular motion of the trawl sack—generates forces which crush the catch and reduce it's quality and possibly it's value.

[46]The word skeppare is constructed from the stem "ship," and (from the grammatical application of the productive morpheme -are) implies simply "one who (as a professional) works with (drives) ships."

[47]This local perception is somewhat at odds with Löfgren's (1978) analysis of the evolution of a fishing community on the coast, well south of Göteborg.  While Löfgren, like myself, also employs a functional framework for explaining social organization, he seems to limit functionality to a notion of "means of production," leading him to underestimate the role of cultural constructions.  Thus, he claims the influence of patrilineality in boat ownership—so dominant throughout fishing societies in the North Atlantic (cf. Andersen and Wadel, 1972)—to be a product of changes in the means of production which entailed the intensification of capital yielding changes in the organization of ownership.  My own data suggest that ownership and crew membership have always (at least as far back as the late 19th century) been closely linked with patrilineality, and while modern economies concentrate this relationship they do not create it independent of the cultural constructions through which the practice operates.  In my analysis, functionality includes all aspects of practice.  Many of these aspects are historical products, some are current or past economic conditions, others are the constraints imposed by the cognitive accomplishment of tasks engendered by the practice.

[48]Actually, the requirement that a non-driving captain be rested when he takes the helm does make his non-driving role a bit different.  Since he must be awake during long periods of driving, he has a warrant to catch some winks at a time when others may be required to forgo sleep in order to get fish sorted and packed.  In practice, the captain's would tend to join the crew and go sleep-deprived in these cases.  (Sometimes, they will stay at the helm where less physical exertion is required until the job requiring manpower is complete and another captain is free to take over.)  Thus, while lack of sleep marks a difference between captains and others, it is not a difference that regular crew members particularly envy, nor a difference which captains capitalize upon to assert their higher status.  In fact, it seemed to me, in this environment where sleep is a highly valuable commodity, that a captain's fatigue carried a symbolic gesture of sacrifice which—while marking one as captain—served as compensation to the crew for his higher status (cf. Barth, 1966).

[49]Most boats' owners include at least one man who is a generation older than everyone else.  This man generally has a son or two who are also part owners.

[50]The one exception to this rule, also reported in the literature to be an old tradition, occurs with first-year fishermen who may—if they are young and financially supported or dependent—only receive a partial share in their first season.

[51]See, for instance, the discussion in Chapter 2 about the short turn-around time for upgrading a ship along this region of the west coast.

[52]It should be noted that not all fishermen find doing less work more desirable.  In fact, general attitudes suggest that the prototypical fisherman is quite eager to work, works hard, and never tires.  More accurately, perhaps, were individuals' reports of boredom associated with trawling many hours (meaning low returns) and with nothing to do but roll with the swells.

[53]It is also the case that men often cook meals at home on the weekends, I noticed.  I do not know whether this feature of the domestic scene has changed under the same time period as the reported changes to cook status on board, but I would predict that it has.

[54]No one wants to suffer the consequences of eating lousy food just to keep a youngster in his place!

[55]Notice that my version leaves open what "costs" and "earnings" for the actor are.  They need not be based in currency or political capital.  Stated in these general terms, the equation is congruent with the claim that human behavior is goal oriented or teleological at some level of description.  The problem, as I will elaborate below, is that these abstract human goals do not determine the organization of human behavior—they are only one input to it.

[56]See Barth (1966) for an attempt to relativize the payoff matrix by grounding the building of reward structures in the dynamics of interaction.  The problem with Barth's model is that the relationship between reward structure and means of interaction remains too abstract.  Agents are assumed to be consulting a ledger for the maximization of personal gain through interaction, but this process has no (or, at least, has only weakly developed) real-world constraints incorporated into the model.  This and the following chapters are meant to address the question, "what are the constraints and resources that inform interaction, and what are the consequences for social organization."  This investigation employs a cognitive analysis of the on board practice to instantiate examples of how interaction and social organization are related in this setting.

[57]The terminology they actually use here is to employ the possessive pronoun "his" when talking about the trawl in the water.  So while they never talk about the trawl in the water being "owned" in the abstract sense I am using here, the division of labor associated with the event is marked by the possessive pronoun in their language.  For instance, in explanation of why the decision to turn around during a draw was made by the other captain, the captain of the boat I was on told me "it is his trawl."