Introduction

The purpose of this paper is to contribute to the sociology of exploration, as part of the sociology of science, by intensive study of a single case: Captain Cook's first voyage to the Pacific. I will assume that "exploration" was science in those days, bringing back data for geography, navigation, biology, and ethnology from places beyond the last outposts of European civilization. (Cook himself was of an empiricist bent. In Tanna he finds a volcano with a peak higher than the peak with the vent, thought by "philosophers" to be impossible: "There seems to be room for philosophical reasoning on these phenomena of nature; but not having any talent that way must content myself with stating the facts as I have found them, and leave the causes to men of more abilities." Nobody here but us country boys. Cook, 1970 [1777] p. 72.)

The literate community of England and Western Europe clearly thought of this as science, and thought of many of the people sent with the expedition as "mathematicians," "astronomers," "botanists," "antiquarians" ( collectors of ethnographic "curiosities"), and, not least, as "explorers" with a distinct intellectual function. One of the scientists on Cook's voyage, Joseph Banks, was later elected as the President of the Royal Society and held the post for many years, something of an autocrat, and was critical in raising the royal gardens at Kew to the best botanical collection in the world.

But my central point is that, just as laboratories have to be created by social processes to maintain modern science, so creating and maintaining an exploring expedition is a social accomplishment. I will outline some of the requirements of creating and maintaining that scientific structure that Cook himself thought he had to achieve, and some that he achieved more or less by accident. My general point is that shipboard life was one main 18^th century version of laboratory life, and just as it is a great accomplishment to keep a laboratory going, so it was a great achievement to keep a ship going. It will turn out that it was especially hard to keep the ship social system in order when it was at anchorage on an inhabited island, while land-based scientific activities were carried out.

The paper deals with the following problems of maintaining a social system beyond the farthest outposts of European civilization (the closest was then at what is now Jakarta [then Batavia] on the island of Java): (1) maintaining the ship as a social and technical system that could sail, locate itself, repair itself, and the like; (2) provisioning and reprovisioning the ship so that the people could live through the voyage, and most of them could come back home and bring the information with them; (3) establishing diplomatic relations with the inhabited islands where they landed, so as to maintain and provision the ship and to reprovision it, as well as to carry out land-based observations; (4) maintaining sufficient unity and discipline among the human complement of the expedition so that all the above functions could be carried out.

The general point is that one could not carry out exploration in the Pacific without maintaining a social system that could make discoveries for periods of many months or years, in a place beyond the reach of flows of resources provided by governments, economies, families, and universities of European civilization. All the civilization one would need had to be on the ship. Cook himself was very self-conscious about the requirements, perhaps our first "experimental functionalist" in the sociology of science. He brought the experiment off, and the social system ended up back in England with a lot of new information. Perhaps the most severe risk to the expedition, as it turned out, was actually the diseases of the farthest outpost, probably both malaria and cholera. The information would have got back anyway, copies of the journals being sent separately from Batavia by civilized commercial ships

Captain Cook rediscovered New Zealand, discovered and documented more thoroughly many Pacific islands, explored the East Coast of Australia and re-found the Torres Strait separating it from New Guinea, while he was still a "commanding Lieutenant." Before setting off on his third voyage, he prepared his journal and the log from the second voyage for publication (Cook 1970 [1777]). In the introduction, he wrote a short essay on the functional requirements of a ship for such a voyage, as illustrated by his preparations for the second voyage (idem, pp. xxiii-xxxvi). In part he wanted to explain why he went in a relatively small merchant ship rather than a more reputable large military ship or a big East India merchant ship. The analysis implicitly described the preparations of his first voyage as well. His sketch of preparations for the voyage suggests a more general analysis of the functional requirements of voyages of exploration in the 18^th century, which in turn suggest what to look for in analyses of 19^th and 20^th century exploratory expeditions.

First, though, it is important to get clearly in mind what the goals of exploration of islands and shores was. The central requirement was that they bring back information about where things were, longitude and latitude accurate enough to bring another ship within sight of a goal. Latitude was easy to estimate quite exactly, but before the age of accurate chronometers one had to be able to tell from observations in the ship what time it was on the ship and simultaneously what time it was in Greenwich. The central device was to observe noon from observing the sun from the ship and to keep track of the time from noon by any method accurate over the short run (either a newly developed chonoemeter or an hourglass), and then to observe the angle between the sun and the moon and compare it with a table of when the sun and moon were that angle apart in Greenwich. The difference in time between that observed on the ship and that estimated for Greenwich gave the longitude.

Sometimes Captain Cook and his "astronomer" took up to nine observations and separate calculations (each of which was quite complicated)in order to take the mean (e. g. the entry for Saturday January 14^th 1769 in Cook 1968 [1893], p. 35, where the mean of nine longitude estimate is 52 minutes, roughly 30 miles at this latitude, off the "modern determination"--an unusually large error for him). In addition to the accurate determination of location, Cook described in great detail various straits, wind directions, currents, harbors (with soundings of depths and bottom [anchoring] descriptions) availability of timber, firewood, fresh water, and provisions, and the attitudes and languages of the natives, useful for navigators. We will come back to timbers, provisions, and natives later.

The difference between "finding" an island or a shore or a strait by landing there or observing it and "discovering" it was to bring back information sufficient for someone else to find it again, a reproducible "finding." And that entailed such information as necessary for maintaining a ship and crew and allowing them to anchor and land, as well as the navigational location information. Whether or not the island or shore was worth finding again determined much of the value of the exploration. In addition, one might discover "wonders" on the land, such as the completely different fauna and flora of Eastern Australia, or the statues on Easter island, or the cannibalism of the Maori in New Zealand and in the Marquesas, or the navigational accomplishments of the Polynesians, which would help sell books about the exploration.

Captain Cook was obliged to ship a copy of his journals and logs back to England as soon as he got a chance, in case he or his ship never made it back. This shows that the primary purpose of the trip was to get the information back; if the ship and its crew got back as well, that was all to the good. But to explore, locate, describe, and get information back, one had to build a social and mechanical system in England that could spend long periods beyond the farthest outposts, could deal with military threats, damage to the ship, replenishing stores, navigational errors. He needed to collect information along the way of the most various kinds.

For example, a difference between observed latitude and longitude and that estimated from the log (which is estimated by the speed of the ship compared with the surface of the water and by the direction of sailing) could indicate whether or not there was a significant current in the ocean; significant currents are more common near land, and so are useful for finding the land. Similarly if the swell caused by a wind from a given direction lasts much longer than the wind itself, that indicates that there is not a large piece of land for some distance in that direction, since long swell is only created over long stretches of water, not over land. Cook was looking especially for continents, so had to estimate the probabilities of finding them in a given direction from clues that could be read at sea. If such information gets back to inform the design of the next exploration, that next exploration is more likely to succeed.

Let us first deal with the ship. Cook says:

In addition much of the repairs to the ship would have to be done far beyond "civilization." We will deal later with the problem of security of the land base when the ship was disabled by being "careened" (laid partly on its side) so as to repair the bottom. In addition the iron equipment of the ship might break and have to be repaired in a forge, and a forge was much safer (for a wooden ship) if it were on the land. The sails often had to be sewed, partly because they simply wore out with time and exposure, but partly also because in unexpectedly stormy seas (especially in the far south of the Pacific and Atlantic, often called "the roaring forties") would put too much strain on them and split them. Wood to replace masts and parts of the ship body had to be found and shaped to the purpose, and masts especially required rather special woods. Caulking to stop leaks and paint for protection had to be replaced.

In short, the maintenance of the ship had to be anticipated. Complete tooling and supplies had to either be shipped with the expedition or found in strange seas. On many of the tropical islands that Cook encountered, the trees near the shores mostly bore fruit, and could not be cut down without bargaining. Without horses, logs from the mountains were not alternatives. If as stores were used, the ship became too light to manage easily, one had to find ballast so it would ride deeper (Cook 1968 [first voyage], pp. 27, 65, 74, 77-78, 114-115, 152, 182-190, 192, 205, and an especially dire emergency repair 389--at this last the forge was broken so they could not make a machine for warping into an anchorage).

The fact that this required planning and foresight during the trip as well is shown by the fact that in an especially dire emergency, Cook had to put into the only place he could reach to repair the ship (near what is now Cooktown in Northeastern Australia; Cook 1986, pp. 274-289). After repairing the ship as best they could, they had to get the ship out of the anchorage. Because they had not been able to choose the anchorage, the strong prevailing winds (from the Southeast.) at that time of year (June) prevented them getting out to sea. They had to wait some considerable time for the correct wind to get out of the harbor (1968 [first voyage]. pp. 290 ff.)

Finding a sheltered harbor for exploration of the land often required sounding (measuring this required smaller boats, because if one finds shallows with a big ship, the knowledge comes too late). Such boats were also used for towing the ship when the wind was insufficient, or when the winds and currents forced the ship toward the shore or reefs. They also were used to carry the ends of hawsers to be tied solidly to something on the land, to "warp" the boat in against a current or contrary wind. And the smaller boats were much more convenient for fishing But then these boats also had to be maintained and overhauled.

Ropes were essential to all the control operations, including managing the sails, anchoring, warping, and they wore out and had to be replaced (Cook, 1968 [first voyage], 74-75, 190--at this last, a hawser fixed on the land broke, which meant that the tension of the wind or current fell entirely on an anchor rope, which had therefore to be cut, so two ropes and an anchor were lost, 200-201, 389).

It is not, of course, a great sociological discovery that if one is going to maintain the crew of an long expedition, one has to feed them. But Cook's problem of subsistence was complicated by the fact that the crew was large, few provisions for large crews were available in the deeps, and sailing so far took a long time.

Cook worried most about the unusual parts of provisioning, and thought of the routine providing of biscuit (or "hardtack") and salt meat only when it was a problem of bulk. In contrast, he paid a good deal of attention to changing the usual naval diet so as to prevent scurvy, a vitamin C deficiency disease, which was known to be a consequence of the usual bulk diet. For example, he paid attention to having a ship with enough unused cargo space to stock enough bread and salt meat, and no doubt bought more when he came into "civilized" ports (he usually failed to mention it, but see e.g. First Voyage, p. 27, on getting provisions and at Rio de Janeiro; when the stores rot or must be dried, he mentions it, First voyage, pp. 76-79 ). He paid more continuous attention to supplies of fresh water, which is heavy enough that it had to be found where one of the ship's boats could get to it, and requires either streams on the land or floating loose ice formed on land so the ice was salt-free. He also collected wood (for cooking as well as repairs) just before setting off for a very long voyage to the high latitudes in the South Pacific where one might not be able to get wood (First Voyage, p. 212) But he paid the most attention to anti-scorbutic strategies: for example, to making sure the officers all ate sauerkraut (pickled cabbage, or kimchee without red pepper) so as to persuade the crew to eat it (First Voyage, p. 59). The overall objective was to keep the officers, scientists, and crew sufficiently healthy when at sea for long periods so that the scientific objectives could be reached, and no doubt for humane and self-preservation purposes as well.

At the time, the reason why some foods were better sources of the "anti-scorbutic" effect, and the fact that it was destroyed by cooking, were not known. Thus there was some evidence that citric fruits had an anti-scorbutic effect, but not that this did not carry over to citrus marmalade. And there was lots of evidence that a long-continued diet of biscuit and salt meat caused scurvy, which tended to disappear rapidly in port where a more varied diet was available. Cook collected multiple suggestions of anti-scorbutic diet elements, and seemed to pursue them all simultaneously and continuously. This involved three major continuing problems: (a) taking along a supply of some diet "supplements" thought to have an anti-scorbutic effect, and as far as possible renewing them at stops beyond "civilization"; (b) getting people, especially the crew, to eat strange foods for preventive (and curative) medicine; and (c) trading for and collecting strange fresh greens and fruits as dietary supplements while anchored in strange, "uncivilized," lands (e.g. for "sellery" greens in New Zealand, First Voyage, pp. 166, 168, 190; fruits and pork in Tahiti, pp. 62-65).

Cook was also took account of the effect of provisions policy on maintaining the unity and morale of the ship. For example he commented that he had the policy that fresh and tasty provisions (obtained by trading, hunting, or fishing) ought to be divided equally between "gentlemen" and the crew in order to maintain crew morale (First Voyage, p. 292; see also ibid., p.170). Many of the incidental items of cargo (hatchets, nails, feathers, etc.) were clearly oriented to having something of value to trade for provisions, as well as for "diplomatic" gifts to authorities on islands where the expedition stopped. Money, and authorities for drafts against the British authorities, were carried for "civilized" ports.

Cook also conceived his scientific task to include reporting on these provisioning problems, since further exploration (as well as trade and conquest) depended on keeping future crews alive and sufficiently healthy to work. For example his essay summarizing the findings about the East Coast of Australia contains a short sub-essay (First Voyage, pp. 317-319) on the availability of provisions along the coast.

Thus the aims of exploration and of coming back with the information required sustained attention to different elements of the provisioning and reprovisioning of the expedition.

Captain Cook, after having had many first encounters with the "locals" of various islands in the Pacific on his first and second voyages, reflected on the combined aggressive defensiveness and willingness to enter into peaceful relations of locals on the island he called Tanna in the New Hebrides:

And after meeting with a male cross-dresser who tried to get members of his crew to come into the woods with him, and the interaction disclosing the sex of the man also to the onlooking natives, at which they exclaim "It's a man! It's a man," showing Cook homosexuality is not a custom among them:

The experience on which these reflections are based is reported in most detail in the logs published on the first volume, which more reports on his experiences for an audience conceived of as the British mariners, rather than edited down for a more popular audience. Cook had, by the time of the editing of his journals for the second voyage, become used to being a world authority on exploration, and was much more inclined to write essays in connection with an incident. Most of the practical matters of diplomacy turn out to depend primarily on spatial arrangements backed by carefully arranged, but local, military superiority.

(1) Local Extraterritorial Spaces: Crew, officer, and scientist activities on land often had to be governed strongly by norms specific to the expedition, rather than adapted to situationally bargained norms with the locals. Sometimes substantial resources would be endangered if such activities were carried out in spaces governed by the locals. Specifically, the scientific work had to be done with fine and valuable instruments, which could not be reproduced by the expedition if they were lost, were portable so easily stolen, were obviously "curiosities" never seen before by the locals, and were made of metals that the locals could not otherwise obtain. But also whenever the ship itself had to be careened for repairs to the hull, it was itself vulnerable, the materials and tools could be stolen, and the military superiority from ships big guns had to be sacrificed, or at least was not so portable. Collecting water put at least the casks and the boat at risk, and fully occupied the crew. Successful trading for a store of local provisions clearly creates a store that could be burglarized.

So for example, when Cook came to Tahiti to observe the transit of Venus (thought to be useful for estimating the distance from the Sun to the Earth), the expedition needed a place on land to observe precisely the times (entering onto the sun's face, and leaving it, with the local time being determined by observation of the sun) and other features of the transit. The location could not be on a ship moved by the waves. Cook found a place good for observation on an uncultivated cape, which therefore did not have claims on it by the locals. There was also an anchorage near the shore there, so the guns of the ship could command the area. He had built a palisade of wood, and stationed sentries so that the English controlled access (except for someone who came over the wall and stole one of the [presumably brass] quadrants, used for observing astronomical angles, measuring noon, and estimating North-South directions (Cook 1968 [1893], p. 68, entry for Tuesday 2^nd May 1769). The forge and oven for repairing the ironwork of the ship was also stationed within the fortress, since all iron was of high value to the locals (pp.62-72).

Cook's crew kidnapped local aristocrat who was in or near the fort on 2^nd May, and held him within it, to get the quadrant returned. Cook dealt with them to make sure the aristocrat free and to "make up" for the threat. Cook fairly regularly took hostages to get restitution for something done by the locals, and eventually apparently got killed because of having used the same strategy in Hawaii. The point here is that without extraterritoriality both on the ship and in the port, neither the observations nor the kidnapping and holding hostage would have been possible. Some days (ibid. pp. 72-73, entry for 12^th May) later a ceremony took place in which the "chief" who manned access at the entrance (Mr. Banks, a botanist and English aristocrat) was "recognized" by a new local group in a ceremony by two women, who stripped, then embraced, the "chief" before the entrance. On the 14^th May (idem.) a young man and a girl in her early teens held what was apparently a defloration ceremony in front of the entrance (Banks and Cook give different estimates of age (Banks younger), whether the girl was sexually experienced or not (Banks yes), and of her willingness (Banks willing); my own bet is on Banks, who was something of a libertine--Cook seems not to understand why libido is so strong). Tahitians marking the boundary between them and extraordinary military power,and secret ceremonies with the sun and stars within the port apparently took some powerful "recognition ceremonies."

The point here is not so much the Tahitians trying to make sense of all this, but the great importance of land boundaries, militarily defended, and with assumed and hopefully recognized extraterritoriality up to and including kidnapping, for the central purpose of the expedition. The diplomacy of science in this situation was overwhelmingly a military and diplomatic problem (see also for related encounters, often less peaceful, ibid. pp. 106, 120, 130-133, 136-38, 158, 163-66, and many places I won't detail here on the second voyage).

Part of the knowledge production purposes of the Cook voyage was collection of botanical and other "natural history" specimens and knowledge. This was principally in the control of Joseph Banks, a rich gentleman with much interest in "collections." The basic notion of what natural history knowledge consisted of was "Linnean," specifying the number of kinds of things in the world, and arranging them into categories by their similarities and differences. Banks in particular was also an "antiquarian," as well as a Linnean botanist and zoologist. That is, he collected curiosities and precious objects that showed the variety of human life, arts, and crafts, as well as oddities of plant and animal life. Unlike Cook, Banks was in a "staff" position in the expedition rather than an executive, and his loyalties were to natural history and "polite learning" (Gascoigne 1994, pp. 119-183), rather than to the navigational and astronomical knowledge that was central to Cook himself.

One consequence of this is that Banks would detach himself from the ship as an enterprise and go collecting and otherwise studying the local environment. When there were people in the environment, he also talked to and negotiated with the inhabitants. Banks had enough interaction in Tahiti to become moderately fluent in the language. Tahitian then turned out to be close enough to the languages of most of the islands the expedition went to, the "Polynesian" islands, so that communication was easily established, with either Banks or a Tahitian as translators.

Banks is reported to have had some trouble with ancient languages at the university, but he had two advantages in learning a good deal of Tahitian. In the first place he detached himself a good deal from the ship and crew, and lived among the Tahitians, apparently including taking lovers (his journals are incomplete on that question). He also participated in some Tahitian rituals. He thus had a good deal of opportunity for ordinary conversation (Gasoigne 1994, pp. 160-176; the Tahitian vocabulary and table of comparison of vocabularies of different islands, Cook 1970 [1777], vol. 2, pp. 317-367, with acknowledgement to Banks).

In addition, he knew a good deal about the local environment, especially the plants and animals, so he had good understanding of the principal objects that formed the daily life of Tahitians. Knowing the semantics of the languages well no doubt helped him learn its syntax and vocabulary. Then it turned out he could speak with the people on many of the small islands the expedition touched and of New Zealand, and could judge quickly and clearly which languages were very different (e.g. in New Caledonia, Cook 1970 [1777], vol. 2, pp. 75-145).

This ethnographic body of knowledge was, by modern standards, very thin (and very bigoted). But it was thought useful to bring back (Cook himself devoted a good deal of time to it in the published versions of his logs, in a way that would be considered nearly plagiarism of Banks nowadays) because it would help future navigators, explorers, colonists, and missionaries, and would be another set of "curiosities" to supplement the items collected.

It also contributed to the speculative theories of race and language. Interbreeding human populations almost always learn to speak a common language, so "racial" and linguistic or dialect boundaries tend, in the long run, to coincide. A mix of evolutionary thought and speculative history of migrations developed to account for the distribution of colors and other "racial" features (especially hair, noses, and skull shapes) of populations, and for the analogies and differences among languages and dialects. They were part of the same body of speculations because of their high correlations (Gascoigne (1994) pp. 119-183. "From Antiquarian to Anthropologist"). .

But from the point of view of this paper, what is most important is that trading and diplomacy of the expedition were both much more successful if one could talk to the people one was dealing with. Reprovisioning when the stay was long, as in Tahiti, New Zealand, and Northeastern Australia was immensely facilitated when military power was the background for mutually beneficial trade (it was not beneficial in Australia, presumably partly because the language was very different from Tahitian). The maintenance of the astronomical station in Tahiti and the reprovisioning and repair of the ship there and on other islands were a good deal easier to arrange when Banks (and some of the other "astronomers" aboard) could speak with the natives of the islands. Even in establishing the military boundaries of "extraterritoriality" of the forts, the landing places for boats, or the ship itself, a verbal translation of why and when the muskets and cannon would be used helped establish political power out of the barrel of a gun. Thus Banks's ethnology was an important part of the maintenance of the social system out beyond the outposts of civilization, though not planned for that purpose.

The capacity of the expedition to get unity and discipline in the crew, so that their efforts could be devoted to the tasks of sailing, the maintenance of the ship, and the work on shore, was the responsibility of the Captain. On shore the crew also had military functions, and these were intimately interdependent with the diplomatic problem of maintaining the extraterritoriality of the spaces used by the expedition. Both of these required a high degree of discipline, under uncertain and constantly changing circumstances.

The Captain, with the surgeon, was also responsible for the maintenance of the health of the crew and officers. Cook believed the core of his own responsibility was maintenance of "dietary discipline," so the officers and crew would use all the various dietary supplements with which he was trying to prevent scurvy (at least some of them worked, and his expeditions had less scurvy than any previous long expedition). Generally speaking there were quite different types of disciplinary problems while the ship was sailing and while the ship was at anchor for some time.

Sailing a ship is a collective enterprise, and many people have to work together in a skilled fashion to bring it off (see generally Stinchcombe 1995, Chapter 3). The highest level of skill in the 18^th century was navigation. It required a lot of astronomical observation, a lot of consultation of navigational tables, and a lot of calculation, as well as keeping track of time, speed, and direction of sailing for intermediate navigation between astronomical fixes of ship position. As mentioned above, longitude was especially difficult to estimate at that time. But the maintenance of trim so the ship would sail well, the repair of ship timbers and hull, the repair of iron implements, the storage and maintenance of the provisions and their preparation for eating, sounding bays and harbors and finding passages and anchorages, all were skilled tasks which had to be coordinated tightly with each other. While there was occasional individual deviance in this sort of work that Cook dealt with in the usual 18^th century draconian fashion (e.g. floggings, Cook, 1968 [1893] First Voyage, pp. 65, 80), the main problem generally was the risk of mutiny.

Being beyond civilization meant being beyond the police and military backup for discipline, so mutiny was possible. The mutiny on the Bounty took place on a ship with a mission rather similar to that of Cook, namely taking botanical specimens, especially breadfruit, from the Pacific to the English Caribbean colonies. The main cause of mutiny in the 18^th century was a food (and water) riot (see brief comments on the relation of provision to ship discipline in Stinchcombe, 1995, Chapter 3). For example, on the Bounty the demand for fresh water to keep the plants alive made the crew go onto short water rations. Cook was exceptionally careful (and exceptionally egalitarian) on the question of provisions, and was self-conscious about its importance for discipline.

At sea all the crew depended on the collective management of the ship, especially in extreme situations (see for example the account of the grounding and search for a harbor near Cooktown Australia, (1994 [1777]) First Voyage, pp. 274-289, and danger near the outer reef shortly afterward, pp. 301-307). In some sense then the unity of the crew at sea and a high level of discipline are to be expected from the community of fate between officers and men. It no doubt helped discipline that Cook seemed to be an exceptionally skilled manager of the ship as a system, very observant of the dangers they faced, and ingenious in getting out of trouble.

The situation was quite different when the ship was at anchor and the main work was ashore, especially when good relations with the inhabitants had been established. Then the temptations of an easy life with good provisions and sex ashore competed with hardship, danger, and celibacy at sea. Some of the most delicate diplomatic work of Cook involved getting the Tahitians to return deserters by kidnapping chiefs and sequestering their canoes. It seems also that sometimes the Tahitians kidnapped members of the crew as well (Cook 1994 [1977) First Voyage, pp. 80, 84). The deserters were confined and flogged when returned (ibid., p. 108). There was less trouble with going native in New Zealand, for it was clearly established that the Maori killed and ate foreigners ("enemies")

But maintaining discipline ashore also involved military-diplomatic "incidents" in which crew members or officers interfered with collective diplomatic objectives. Cook arranged for trading for provisions in Tahiti to be centralized under his authority so as to establish regular prices (idem, pp. 60-61), sought out and punished (2 dozen lashes) a man who stole nails for trading individually, and punished a man in New Zealand for digging up potatoes in a Maori garden without permission of the owners (idem, p. 166), Sometimes overreaction to perceived threats or to stealing by the natives created problems (ibid. p. 134; Lyte (1980) pp. 93-98).

Cook enjoyed the monopoly of legitimate violence as an aspect of his job as manager of the ship's discipline, and as the ultimate authority over diplomacy with the inhabitants of the islands and of Australia. There is every evidence that he was conscious of that monopoly, and that he tried to use in rationally and sparingly to achieve the ends of the expedition. He used his monopoly of violence to create a government that usually worked. He was killed on the third voyage in Hawaii when that government failed, and the Hawaiians killed him.

I have tried to show that Captain Cook undertook a self-conscious program to build a social system that could explore for long periods beyond the last outposts of civilization, could recover from disasters and hostile encounters by itself, and could carry out systematic scientific observations, and some not so systematic ones as well. He knew this required maintaining a ship in extreme circumstances, provisioning and reprovisioning the ship so it could maintain the strength and health of the crew and officers, establishing diplomatic relationships with the inhabitants of the places he explored or where he had to beach and repair his ship, and creating a sufficient ethnography of those peoples so that negotiations could go on, and so that they could bring back ethnographic information as well. This is a social system of a complexity of the same order of magnitude of our great physics and astronomy labs, and it had to be pretty much self-sufficient because the resources of civilization had to come with them. I suppose its nearest modern analogy is a space station.

I urge on the reader that looking at the information that came back with such exploratory expeditions through the lens of what kind of social system had to be produced to get that information throws new light on 18^th century science.