Figure 1: Khizhin’s characteristic curves (from average data) for amount of gastric juice secretion (in cubic centimeters) in response to various foods. Khizhin, Otdelitel’naia, appendix, based on average data on pp. 65, 71, 88, 93
Pavlov and his coworkers experimented with many scores of dogs, but these were not created equal. Some possessed physiological and temperamental attributes that made them, in Claude Bernard’s phrase, a “happy choice” of experimental animal. They varied greatly in their cooperativeness on the experimental stand, excitability and food tastes, and ability to generate pravil’nye results. Dog technologies were also endowed by their laboratory creators with different surgical modifications and occupied different places in the development of the lab’s lines of investigation.
Pavlov’s scientific style would always feature the intensive analysis of such favored individuals. These “template dogs” can be identified by name for almost every major development in his research on digestion and conditional reflexes from the 1890s through 1936.
Temperament, technology, and timing combined to make Druzhok the template dog for the gastric glands, Zhuchka for the pancreas. Under the chief’s supervision, three coworkers—Pavel Khizhin in 1894, Ivan Lobasov in 1895–1896, and Andrei Volkovich in 1897–1898—studied the gastric glands through experiments on Druzhok, the first dog to perform successfully with a Pavlov sac. Their counterparts for the pancreatic gland were Anton Val’ter and Zhuchka, the first dog to do so with the troublesome pancreatic fistula, in experiments of 1896–1897. These dogs’ secretory responses became the standard against which those of other dogs were assessed, and were enshrined in what Pavlov termed “characteristic secretory curves,” which greeted readers of his synthetic work, Lectures on the Work of the Main Digestive Glands, illustrating dramatically his central argument—that the digestive system was a precise and purposive “chemical factory.”
Pavlov’s path from dog to digestive factory, then, emerges clearly in the lab careers of Druzhok and Zhuchka. Here the creative tension at the heart of his scientific style is evident: his attempt both to understand the intact, normal, functioning organism in all its complexity and to find there fully determined, precise, and, ideally, quantitative laws. This tension lent an inherent interpretive dimension to even the most apparently esoteric experiments, to say nothing of his efforts to deploy them to prove his central argument. It would also prove a main source of his appeal as a scientist who bridged the gap between experimental physiology and clinical medicine.
The creation of a dog with a Pavlov isolated sac flowed from Pavlov’s earlier identification of the two central agents of gastric secretion and his desire to investigate their joint action in an intact and normal animal. Using the esophagotomized dog that he had devised in 1889–1890 and improved in his new Physiology Division during the early 1890s, he had demonstrated that the first phase in gastric secretion was controlled by the psyche and elicited by appetite, by “a more or less lively representation of food.” Sanotskii and Pavlov had also discovered that the qualities of this secretion—its quantity and proteolytic power (its ability to break down proteins)—differed from dog to dog and day to day. It was governed, in other words, by the individual psychological qualities and mood of the experimental animal. Experiments upon dogs equipped with Heidenhain’s isolated sac had also confirmed that the presence of food in the stomach produced a second, weaker gastric secretion—a “nervous-chemical” phase in digestion.1
For Pavlov, then, there existed two mechanisms of gastric secretion: a “distinctive or idiosyncratic psychic process” transmitted through the vagus nerves and generating a “very active” product (juice with a high pepsin content); and a second mechanism, “also nervous,” that was excited by the presence of food in the stomach and produced a much smaller and weaker secretion.2
This second phase could not, of course, be investigated with esophagotomized dogs (since the food they swallowed never reached the stomach), nor, in Pavlov’s view, through dogs with Heidenhain’s isolated sac, since that sac lacked normal vagal innervation. The task, then, was to create an isolated stomach with intact vagal nerves.
In the fall of 1893, Pavlov collaborated on this signal surgical feat with the Ol’denburgskiis’ talented physician and surgeon Pavel Khizhin. At least sixteen dogs perished under the knife or shortly thereafter, and two more survived for only a couple of weeks. Shortly thereafter, the director of the IEM fell ill and Pavlov temporarily assumed his duties. Burdened with unfamiliar administrative tasks and assured by local pathologists that the operation was doomed to failure, he considered abandoning the procedure while the panic-stricken Khizhin, his leave time drawing short without a completed dissertation in sight, despaired of ever receiving his doctorate.3
Finally, on April 2, 1894, nature and technique combined to produce what the relieved experimenters christened “the remarkable Druzhok” (Little Friend). Khizhin apparently performed the successful operation while Pavlov was mired in administrative duties—though he did so, of course, according to Pavlov’s conception. Four days after the operation, Khizhin reported that Druzhok was “energetic and happy.” By April 9 his appetite was “excellent,” and on April 13 he took his first postoperative stroll around the grounds. The next day there began five months of experiments, all conducted while the dog was in “excellent health.”4
Druzhok’s isolated sac, combined with the now-doctrinal view of the importance of the psyche, made this dog not a mere site of glandular secretion, but rather an active subject whose character demanded the attention of experimenters. Druzhok confronted experimenters with the complex relationship between psyche and glandular response, and between the psychic and nervous-chemical agents of secretion.
Work with Druzhok required considerable patience. Most experiments lasted about five hours, and some as many as ten, during which Khizhin strove to avoid exciting the animal with movements or sounds. Having ascertained that the animal’s gastric glands were at rest, he then teased or fed Druzhok, waited about five minutes for the first drops of gastric juice to appear in the special fistula that ran from the isolated sac, and collected the secretions at fifteen-minute intervals. When the animal was merely teased with food, secretion continued for only about two hours, but when Khizhin fed the dog milk, bread, meat, or mixed food, he had to remain as still as possible for five to ten hours at a stretch.5
Druzhok, too, needed to be, as Pavlov observed in an homage to the experimental dog, “understanding and compliant.” The animal’s willingness to lie quietly on a table during the long trials greatly facilitated the experiment’s success. Otherwise, given the length of these trials, both investigator and dog would tire, undermining the precision of experimental results. For one thing, Khizhin noted, the dog’s exhaustion “could hardly fail to be reflected in the course of secretory activity.” For another, an “exhausted or simply bored animal” would inevitably make jerky movements that disrupted the collection of gastric secretions. Given the small quantity of secretions from the isolated sac (the surface area of which was estimated to be only 10 to 20 percent of that of the intact stomach), failure to catch even small quantities in the collecting cup could badly skew the results. Druzhok adapted rapidly to this requirement, lying peacefully on the table and “taking no particular interest in anything” during the experiment. Better still, he frequently slept for hours at a stretch.6
Although Khizhin attested to Druzhok’s excellent health, he knew that one problem with the isolated sac raised a question about the dog’s normalcy. Gastric juice flowing from the aperture of the isolated sac digested the surrounding tissue, causing a sinusoid ulcer and considerable pain. This and other questions about the dog’s health would serve throughout Druzhok’s career as a source of interpretive challenges and flexibility.7
In April and May 1894, while Pavlov tended to administrative duties, Khizhin conducted experiments assigned by the chief to compare the functioning of the Pavlov and Heidenhain sacs. A series of “characteristic particularities” demonstrated the superiority of Pavlov’s variant—it produced a psychic secretion, had a shorter “latency period” between stimulus and response, and yielded gastric secretion of higher proteolytic power. It offered, Khizhin concluded, “a pure and true mirror of the stomach in which one can with complete clarity observe the activity of this organ in all its smallest details.”
Another difference was especially intriguing: the quantity and quality of gastric secretion in the Pavlov sac, unlike that in Heidenhain’s version, varied independently, “each pursuing its own goals.” That is, the amount of gastric juice and its proteolytic power did not rise and fall together. For Pavlov, the regularity of this phenomenon attested to the existence of a determinist relationship with a particular purpose, and so to the existence of a specialized nervous mechanism.8
On June 2, 1894, just as Khizhin was completing his experiments on the qualities of Druzhok’s new stomach, Pavlov completed his tenure as acting director of the Institute. Three days later, he first mentioned Khizhin’s research in a letter to Serafima, who was summering with their son in the countryside: “I am now delighting in total concentration on Khizhin’s experiments. An enlivening of our projects is inevitable—success after success, not only new but downright beautiful.” Two days later he postponed a scheduled visit to his family: “Khizhin’s work is flowing so successfully, and is of such gripping interest, that I want to see it all myself.”9
The encouraging performance of Druzhok’s new stomach certainly promised “an enlivening of our projects,” but what exactly did Pavlov find “not only new but downright beautiful”? Khizhin had just embarked on a series of trials to test Druzhok’s secretory response to various substances that, when present in the stomach, might reasonably be expected to excite gastric secretion. In an attempt to avoid arousing the dog’s appetite—and so eliciting a psychic secretion that would obscure the nervous-chemical effects of these substances themselves—he used a cannula (a long, thin hollow tube) to introduce material directly through the throat into the stomach. In this manner, he tested Druzhok’s response to water, acids, alkalis, salts, starch, egg white, and a commercial peptone prepared by the Parisian pharmaceutical factory Chapoteaut.
Khizhin and Pavlov had good reason to believe that peptone excited the gastric glands. Sanotskii’s experiments had revealed that the first, psychic phase in gastric secretion produced a large quantity of pepsin-rich juice, which greeted the food mass as it arrived in the stomach. The experimenters reasoned that the substance that excited the peripheral nerves of the stomach membrane—initiating the second, nervous-chemical phase of secretion—must be either a common component of food or a product of the contact between food and the pepsin contained in psychic secretion. They reasoned further that a dog’s usual meal of mixed food contains albuminous substances that might be converted to peptone by the pepsin in psychic secretion. Therefore this peptone might well be the exciter of the second phase of secretion, in which case peptone placed directly in the stomach should itself excite the gastric glands.10
Pavlov’s excited letter to Serafima referred to experiments that seemed to dramatically confirm this hypothesis. In each trial, the introduction into Druzhok’s stomach of a solution of Chapoteaut peptone resulted in an abundant secretion. The dramatic excitatory effect of peptone manifested striking similarities in each trial. Not only the total secretion but also the data for hourly and even fifteen-minute periods were, as Khizhin reported, “remarkably similar and almost identical, like clockwork.”11 It was this feature of Khizhin’s trials—this precise, repeated, mechanical regularity of secretory responses to a specific exciter—that seemed to Pavlov “downright beautiful.” And these wonderfully pravil’nye results promised to be only the beginning. On June 4, Khizhin began his first experiments on Druzhok’s secretory response to feedings with raw meat—a signally important development in the direction of experiments (“an enlivening of our projects”). Previously, Khizhin had fed Druzhok mixed foods to test the functioning of the isolated sac. But the meat experiments that began on June 4—and which, together with similar experiments with milk, consumed the entire month of July—were designed to compare patterns of secretion for different foods. The search for distinct secretory patterns during the digestion of different foods, and the attempt to describe these patterns in undeviating, standard secretory curves, became a central theme of Khizhin’s thesis and of Pavlovian digestive physiology.
An epilogue to these peptone experiments highlights the omnipresent role of interpretation. The results that delighted Khizhin and Pavlov in June 1894 were discarded one year later as meaningless artifacts. Based on a series of experiments in 1895, Pavlov and Ivan Lobasov, the next coworker to experiment upon Druzhok, concluded that peptone did not, in fact, excite gastric secretion. Khizhin’s results were discarded—and attributed to impurities in Chapoteaut peptone and to the inadvertent excitation of Druzhok’s appetite.
By this time, however, Pavlov had already glimpsed the outlines of a secretory apparatus that produced precise, repeatable, and distinctive responses to different exciters, and the resultant experiments had already produced much confirming evidence. In his Lectures of 1897, he publicly acknowledged his earlier error about peptone; but those “downright beautiful” results had already given rise to a view—one resonant with Pavlov’s deepest convictions—that careful, methodologically sound experiments on the gastric responses to various foods would discover distinctive, stable, and determined secretory patterns—“an exact scientific formula” for a complex physiological process in an intact animal.12
Khizhin’s experiments on Druzhok during the summer of 1894 yielded the “characteristic secretory curves” that became the logical and rhetorical centerpiece of Pavlov’s digestive physiology, dramatizing the ability of the gastric glands to respond to various types of food “precisely and purposively,” producing secretions of precisely the amount and fermenting power necessary to digest them. These curves represented Pavlov’s identification of meaningful patterns in experimental data. Like the normalcy of experimental animals and the roles of the psychic and nervous-chemical mechanisms, they were presented as simple empirical facts but inevitably represented a series of interpretive and rhetorical decisions—and these, in turn, changed with changes in lab doctrine.
The curves originated in the data from a relatively small number of experiments on Druzhok. In his doctoral thesis, Khizhin reports that he experimented on the dog’s response to normal feeding in the following trials: five experiments each with 100, 200, and 400 grams of meat; five experiments with 200 grams of bread; and three with 600 cubic centimeters of milk.13 He attempted only three trials with milk because Druzhok disliked it, refusing it altogether after his first encounter with the nonfat variety. Conversely, the conduct and interpretation of trials with bread were complicated by the animal’s passionate preference for it. For all these reasons, in his thesis Khizhin identified these characteristic secretory curves with some tentativeness.
Even identical trials with the same quantity of the same food yielded varying—sometimes widely varying—secretory results. Pavlov and Khizhin were fully aware that the results of any trial were influenced by Druzhok’s personality and mood, yet they sought to find in the quantitative results of their trials a pravil’nyi pattern. This necessarily involved judgments about the relative importance or unimportance of similarities and differences between the data generated by various trials, and about the reasons for those differences.
Physiology at this time offered no established traditions for making such judgments. Physiologists were largely unacquainted with the development of statistics and, in any case, had yet to negotiate the tension between statistical thinking and the determinist ideal enunciated by leading figures in their discipline. (Bernard, for example, had famously rejected statistical reasoning as a cover for the failure to find truly determined relations.) As Pavlov sought to make sense of his varied data, he was compelled, like others in the same situation, to develop his own, homegrown approach.14
That approach was basically this: the data from “good experiments” (those judged to be undistorted by uncontrolled variables) were averaged, these averages were used to construct a curve, and that curve was illustrated with the results of a single model experiment. This procedure minimized the importance of varied results within any single category—between various experiments for, say, the proteolytic power of gastric juice elicited by a meal of meat—and dramatized differences between categories. By constructing curves on the basis of average data and then presenting them in the form of a model experiment, this procedure worked rhetorically to bridge the gap between Pavlov’s resort to average data (which was necessitated by the varied results of individual trials) and his determinist ideal. For those unfamiliar with the intricacies of curve construction, this practice created the impression that the results of every experiment basically conformed to those in Pavlov’s choice of model experiment. But this was hardly the case.
The general interpretive issues inherent to curve construction are clear if we ask the same question that Khizhin and Pavlov asked themselves in 1894: which of the curves in Figures 1 and 2 are “essentially” the same, and which are different?
Figure 1: Khizhin’s characteristic curves (from average data) for amount of gastric juice secretion (in cubic centimeters) in response to various foods. Khizhin, Otdelitel’naia, appendix, based on average data on pp. 65, 71, 88, 93
Figure 2: Khizhin’s characteristic curves (from average data) for proteolytic power of gastric juice (in millimeters, as measured by the Mett method) in response to various foods. From Khizhin, Otdelitel’naia, appendix, based on average data on pp. 65, 78, 88, 93
There is an ineffably subjective dimension to this task. Especially in the absence of a single, conventional statistical methodology, one can choose a number of plausible groupings (or none at all) depending on which aspects of the curves one decides are important, how much “stretch” one is willing to grant within a category, how little stretch one allows between categories, and how much confidence one has in the data.15 In Pavlov’s lab, such interpretive moments were always shaped by prevailing doctrine.
For Khizhin and Pavlov in 1894, all the curves in Figure 1 were essentially the same, while those in Figure 2 fell into four subgroups: a, b, and c; d; e; and f. In other words, the curves for the amount of gastric secretion (Figure 1) were considered the same for various quantities of meat, milk, and bread. The total volume of juice differed for each food, but each secretory curve exhibited “one and the same regular course, which can be expressed in the form of a curve that attains its acme either during the first hour of the digestive act (the ‘psychic type’) or only in the second and third hour (the ‘usual type’).” That is, when Druzhok responded enthusiastically to a meal, the resultant “psychic type” of secretion reached its peak in the first hour, reflecting the large flow of “appetite juice”; when he did not, and so produced little or no appetite juice, he produced the “usual type” of secretory curve. In each case, immediately after reaching its peak, this standard curve began to fall. The differences between these curves were considered secondary.16
The curves for proteolytic power (Figure 2), on the other hand, were considered to be fundamentally different, expressing “sharply expressed distinctions” in the glandular response to different foods. “To each type of food,” Khizhin explained, “there corresponds its completely special [level and] course of the proteolytic power of juice during the digestive act; this course is in each case so different from that for other foods, and is repeated with such consistency...that we are compelled to consider it characteristic for the given type of food.” The curve for raw meat declines from the very beginning of the digestive act, followed by a rise to a point surpassing its initial level. The “special character” of the bread curve resided in its very high level in the first hour, its further rise during the second, a plateau throughout the third and sometimes the fourth hours, and a decline of about 15 percent in the fourth and fifth hours. The milk curve was striking because it was “completely opposite” to that for bread: beginning at a relatively high level, it fell over the next hour to one-half its previous value, remained at that lower level for the third and fourth hours, regained its previous height in the fifth hour, and maintained that level even in the sixth hour.17
Khizhin appended the characteristic curves to his thesis and invoked them constantly to help his readers properly “distinguish the primary from the secondary.” In his report to the Society of Russian Physicians, he provided a visual display of these curves to convince his audience about his groupings. In the edited form of his thesis that appeared in the French edition of the Institute journal, Archive of the Biological Sciences, the curves were moved from the appendix to the text better to deploy their rhetorical power. Pavlov would do the same to great effect in his Lectures on the Work of the Main Digestive Glands.18
Khizhin’s experiments with Druzhok also underscored the central role of the dog’s psyche and temperament. The secretory results of teasing experiments, he discovered, “depend to a great degree on the individuality of the animal, and also on some other causes which give to this very individuality rather broad fluctuations.”19 To conduct meaningful experiments, then, one needed to understand and adjust to Druzhok’s personality. Khizhin wrote the following, for example, about experiments to elicit a psychic secretion by teasing Druzhok:
In view of the fact that, as he proved more than once, Druzhok possessed an unusual impressionability and a broad self-esteem, we needed to approach this teasing with a special delicacy; otherwise—as actually occurred—having noticed that we are only teasing him (by crudely and immediately snatching a piece [of food] away when he reaches for it)—he turns his snout away and does not even want to look at the things around him. Therefore, in order to attain our goal and interest Druzhok with the teasing, we would carry in a plate with meat, milk, and bread and place it near his snout, avoiding even the appearance that we wished to tease the dog; we would cook a piece of it on the gas flame, pour the milk, and cut the bread; under such delicate conditions, the dog immediately took an interest in our activities and began to get disturbed—inhaling energetically, stretching out its snout in order to draw closer to the food, began to intensely smack its lips, to swallow from the plentifulness of saliva; at the very same time, juice appeared in the tube introduced into the aperture of the isolated stomach.20
As Pavlov put it, if the dog “guesses” it is being deceived, teasing will not produce a psychic secretion. “The dog is an intelligent animal and is angered by this ruse no less quickly than a person would be.”21
Seeking to distinguish between the psychic and nervous-chemical components of Druzhok’s secretory responses, Khizhin attempted to circumvent the psyche by introducing foods directly into the dog’s stomach through a cannula. The results proved puzzling. For example, Druzhok’s secretory response was greater when milk was introduced directly into the dog’s stomach than when ingested normally. This seemed to contradict lab doctrine, which regarded total secretion as the sum of psychic secretion and nervous-chemical secretion. If A (psychic secretion) + B (nervous-chemical secretion) = C (total secretion), how could B be greater than C? Khizhin’s (and Pavlov’s) interpretive response exemplifies the use of the psyche as a flexible explanatory variable. Perhaps, Khizhin reasoned, Druzhok responded with a psychic secretion even to feeding with a cannula: a drop of milk usually fell on the dog’s tongue during this procedure, and even when it did not, the dog’s “extraordinarily subtle sense of smell” virtually guaranteed a psychic effect. Perhaps, since the animal disliked milk, the prolonged, awkward, and disturbing activities associated with shoving a long tube down the animal’s throat produced a stronger psychic reaction than did feeding it this relatively unappetizing food.
In any case, since Druzhok could not be fed without noticing, all conclusions about the secretory effects of nervous-chemical factors remained tentative. “The psychic excitation of the animal has perhaps not been entirely excluded,” Khizhin acknowledged, “and consequently, we must be very cautious in our conclusions.” 22
This admission condemned Druzhok to further operations. In September 1894—while Khizhin hurriedly composed his doctoral thesis and his successor, Lobasov, waited in the wings—the large stomach of the lab’s prize dog was fitted with a gastric fistula. Lobasov could now test the nervous-chemical response to the presence of various foods in the stomach by feeding Druzhok directly through the fistula, bypassing his mouth (and presumably his psyche) altogether. Toward the end of Lobasov’s experiments, in April 1896, Druzhok underwent two more surgical procedures: an esophagotomy and fistula were added to permit experiments on the dog’s secretory response to sham-feeding. These additions reflected Pavlov’s determination to separate the different mechanisms of digestive secretion—that is, to study the psychic and nervous-chemical phases in isolation.
* * *
In December 1894, Pavlov addressed the Society of Russian Physicians about his lab’s findings and their implications for medicine. Emboldened by Khizhin’s successes with Druzhok, he spoke most confidently. Invoking years of research and the findings of sixteen coworkers, he presented a systematic analysis of the digestive system and the relationship between experimental physiology and medicine. He was fairly bursting with results, research perspectives, and general observations, and his speech proved too long for a single Society session.
In this speech, Pavlov introduced the central idea that now imparted a general direction to research and interpretation in his lab: the digestive system was a “complex chemical factory” that responded precisely to the requirements for processing various foods.
This metaphor apparently emerged toward the end of Khizhin’s labors in mid-fall 1894. Earlier lab publications referred to the “phenomena” or “exciters” of digestive secretions, but a new language emerges in Khizhin’s general introduction, the conclusion to one critical section, the general conclusion, and the title of his dissertation. Here, on pages no doubt written or rewritten in consultation with Pavlov, Khizhin refers to secretory “work” (rabota), as in the title of his thesis, The Secretory Work of the Stomach of the Dog. From this time forward, the word “work” entered the lexicon of lab publications, most famously in Pavlov’s Lectures on the Work of the Main Digestive Glands.23
This word represented the tip of the factory metaphor that Pavlov introduced in his speech of December 1894:
The digestive canal is in its task a complex chemical factory [zavod]. The raw material passes through a long series of institutions in which it is subjected to certain mechanical and, mainly, chemical processing, and then, through innumerable side streets, it is brought into the depot of the body. Aside from this basic series of institutions, along which the raw material moves, there is a series of lateral chemical manufactories [fabriki], which prepare certain reagents for the appropriate processing of the raw material.
A series of research questions flowed from this conception:
What is the activity of this factory at full operation, how and by what is it brought into motion, in what manner does one part go into operation after another, in what manner does the work change in dependence upon the type of raw material, does the entire factory always operate with all its parts, or not?.... One cannot doubt that in the investigation of this subject we will find the very same subtlety and adaptiveness of work that strike us in other, better-studied areas of physiology.24
Like so many metaphors in the history of science—Darwin’s “struggle for existence” and our own “genetic code,” among others—Pavlov’s factory metaphor was not simply a rhetorical device, but rather a way of thinking. It offered an approach to a complex phenomenon that he was seeking to understand (the digestive system) by reference to another (the factory) that he thought he already understood. This metaphor had three general implications for lab research and interpretations.
First, the factory metaphor placed an emphasis on the digestive apparatus as a precisely coordinated system operating toward a single end. This perspective defined a series of questions about the operation of that system. The subtle and adaptive work of the digestive factory required an equally subtle coordinating mechanism. For Pavlov, the only conceivably satisfactory mechanism was the nervous system with its property of specific excitation. In its role as coordinator of the digestive factory, the nervous system, then, came to bear increasing explanatory weight as the often confusing, even paradoxical data from chronic experiments piled up. To accommodate this data, the nervous system acquired a number of hypothesized structures and properties.
Second, the factory metaphor expressed Pavlov’s determinism in a new, refined, exacting manner, channeling the search for pravil’nye results in specific directions. The factory metaphor focused attention on the specific purposive, determined relations to be uncovered experimentally: that between the raw material and its processing—that is, between the ingested food and glandular responses to it. This, in turn, put “pressure” on experimental data, encouraging the quest for determined and precise secretory patterns.
The search for precise, “factory-like” patterns in the varying data generated by chronic experiments gave rise to a tension between Bernardian principles and Pavlov’s experimental practices. Bernard had firmly rejected the use of statistical techniques as incompatible with the physiologist’s determinist creed. In Bernard’s narratives about scientific method, experimental trials always ended with a definitive, repeatable conclusion that reconciled earlier “positive” and “negative” results. These narratives, however, always involved questions that could be answered with a simple yes or no. (For example, does piercing a rabbit’s brain with a needle cause an increase in the sugar content of its blood? Sometimes it did and sometimes it did not—but rather than settle for a meaningless statistical statement, Bernard related, he finally determined that the result depended upon specific variables—the location and depth of the piercing.) Pavlov’s factory metaphor pressed him to provide similarly definitive responses, but to a very different kind of question, and, as we have seen, using a very different kind of data (the averages from experiments that always differed somewhat in their results).
The factory metaphor also highlighted the contradictory relationship between the idiosyncratic psyche and the pravil’nyi digestive machine. Previously, Pavlov could be true to his determinist views by simply recognizing the psyche’s important role in shaping glandular responses to food and interpreting experimental data accordingly. There was no need to confront the nature of the psyche, which was simply “black-boxed” as an important, poorly understood factor that might even lie outside the bounds of determinist physiology. The factory metaphor, however, incorporated the psyche fully within the digestive machine, transforming the long-appreciated relationship between that machine and the idiosyncratic psyche into a contradiction. The acknowledged importance of the psyche remained a source of interpretive flexibility in experimental trials, but an uneasy dualism developed—not between mind and body per se, but between capriciousness and pravil’nost’. How could a factory haunted by an eccentric ghost be purposive, precise, and determined?
The appeal and meaning of this metaphor for Pavlov resided in his social context and personal attitudes. The image of factory production permeated late-nineteenth-century discourse, and was omnipresent in Russia during its industrial revolution of the late 1880s and 1890s. Yet Russian intellectuals regarded factories very differently. Many, on both the monarchist right and the populist left, saw the emergence of huge factories in St. Petersburg and other large cities as an ominous sign of Western-style capitalism with its attendant evils (most commonly, the emergence of an urban proletariat, class struggle, and the destruction of cottage industry and Russia’s communal social fabric). From his days as a youthful admirer of Pisarev and Smiles, Pavlov, however, had a Westernizing bent, and instead associated factories with a set of positive attributes: with precise, powerful, efficient, effectively coordinated, modern production for a particular goal.
His personal psychology—his dread of sluchainost’ (accidents, unforeseeable events) and the great value he attached to personal control, discipline, integration, and efficient productivity—doubtless also played a role here. The factory metaphor may well also have gained commonsensical power for Pavlov from his own enterprise, which he strove to make purposive, regular, and precise. Furthermore, by setting before Pavlov a panoramic view of results on different glands, his lab afforded him a view of digestive mechanisms as a system. So, for example, whereas Khizhin had demonstrated (at least temporarily) that peptone was the specific exciter of the gastric glands, another coworker, Ivan Dolinskii, had discovered that the hydrochloric acid produced by the gastric glands served, in turn, as an exciter of the pancreatic gland. This made “the joint activity of various parts of the chemical factory” much more striking than it would have been had Pavlov been a lone investigator addressing one gland at a time.
Pavlov apparently had no direct knowledge of factories, but his longtime acquaintance Dmitrii Mendeleev did. Best known in the West for his periodic table of the elements, Mendeleev also served as chief of Russia’s Bureau of Weights and Measures, and was a leading authority on economic, industrial, and technological developments. In that capacity, he wrote the article “Factories” in the twelfth volume of the authoritative Brockhaus and Efron encyclopedia, which appeared in 1894—on the eve of Pavlov’s landmark speech to the Society of Russian Physicians. Several important nuances common to Mendeleev’s article and Pavlov’s speech suggest that Pavlov may have drawn upon the article (or, perhaps, on a personal conversation) in his conceptualization of the digestive factory.
In his article, Mendeleev emphasized the distinction between manufactories (fabriki), in which raw material was subjected to primarily mechanical alterations, and factories (zavody), which relied primarily on chemical processes that produced more profound “molecular transformations of substances.” The latter, Mendeleev emphasized, were the more advanced, making chemical factories the cutting edge of industrial progress. This same distinction is evident in the passage from Pavlov’s 1894 address in which he describes the glands themselves as “manufactories” and the digestive system as a whole as a “factory.” (The glands simply produce digestive juice, whereas the actual chemical alteration of food occurs in the digestive canal.) Pavlov further noted that the digestive canal is a “complex chemical factory” in which raw material “is subjected to certain mechanical and, mainly, chemical processing.” By Mendeleev’s criterion, then, the digestive system matched the most advanced type of factory in existence.
Mendeleev also observed that modern factory production differed from craft production, not only by the number of workers involved, but also by “the presence of particular [specialized] knowledge demanding preliminary preparation, and also of machines and apparatuses that act, although under the guidance of people, mainly by means of the forces and phenomena of nature [as harnessed by science and technology],...for example, by means of furnaces, steam machines, and so forth.” For Pavlov, the controlling apparatus in the digestive factory, and the repository of its “specialized knowledge” (in the form of specific excitability), was the nervous system. The precision of glandular processes, he asserted in 1894, was “of course possible only through the participation of the nervous system, of this regulator, coordinator, of the activity of various organs.”25
Pavlov’s notions about the factory were based not on any actual experiences, but entirely on an idealized image. The factories of the time were rarely so regular and precise as he imagined; their workers and managers struggled with the same difficulty in obtaining regular and reliable results as did Pavlov and his coworkers. Yet it was Pavlov’s metaphorical, ideal image of the factory that guided him conceptually and that he glimpsed in the data generated in his lab. For him, only the inevitable uncontrolled variables in physiological experiments prevented them from yielding the same results with recognizably “factory-like” precision and regularity.
The same passages of Khizhin’s thesis that employed the word work also introduced another term related to the factory metaphor: digestion was “never accomplished according to any shablon.” A shablon was a template used to produce the same shape time after time. Khizhin’s point was that the secretory curves produced by different foods were not produced by a single template; rather, they differed for different foodstuffs. In his thesis, he described the work of the digestive glands most frequently (seven times) with the chief’s favorite word, pravil’nyi, and twice as proceeding zakonno (lawfully).26
Khizhin (and Pavlov) saw these qualities as expressing a deeper truth about glandular work: its purposiveness or, in more contemporary biological language, its adaptation to the requirements for efficiently digesting any meal. This interpretive framework expressed Pavlov’s belief that the organism was perfectly fitted to its environment, and that its parts existed in perfect harmony one with the other—views rooted in Bernardian physiology and in Russia’s evolutionist tradition.27
It is an interesting reflection of the division of intellectual property in the lab that, although the factory metaphor structured the design and interpretation of experiments, and the language in which results were reported, this metaphor was never expressed directly in coworkers’ reports and articles. This all-encompassing expression of the lab’s basic view, which united experiments and gave them meaning, belonged to Pavlov, and was reserved for his statements of the grand vision.
* * *
Ivan Lobasov inherited Druzhok from Khizhin in 1895, by which time the factory metaphor was organizing Pavlov’s ideas about the digestive system. The next turn in Druzhok’s lab career, Lobasov’s charge from the chief, and the “hardening” of Khizhin’s tentative secretory curves all flowed from that metaphor’s implications.
Lobasov was assigned the task of more fully incorporating the idiosyncratic psyche into the pravil’nyi digestive factory, a factory now believed to generate distinctive secretory responses to different foods. His charge from Pavlov was this: “To demonstrate the consistency of [psychic secretion], to detect and distinguish in each case of the secretory work of the stomach during feeding that which relates to psychic secretion, and to elucidate to what degree the typicality of secretion with various sorts of food is determined by the participation of the psychic moment.” The digestive factory could not work regularly, precisely, and purposively if the most important force in its operation, the psyche, was entirely capricious. Psychic secretion, too, must necessarily manifest, as Lobasov put it, some “consistency.”28
Druzhok had been re-equipped for this research by the implantation of an esophagotomy and gastric fistula. The fistula would allow Lobasov to feed him while bypassing the animal’s eyes and mouth (and so, presumably, the psyche), and the esophagotomy would allow him to test Druzhok’s secretory response to the act of eating alone—in other words, to analyze psychic secretion in isolation from nervous-chemical mechanisms.
The effort to incorporate the psyche more fully within the digestive machine is evident in Lobasov’s reinterpretation of the characteristic secretory curves. For Khizhin (and Pavlov) in 1894, the course of the secretory responses to various foods differed with respect to proteolytic power, but not with respect to quantity. The latter expressed “one and the same pravil’nyi course,” differing only in that some (the psychic type) reached their acme during the first hour whereas others (the usual type) peaked in the second hour. For Lobasov and Pavlov in 1896, however, these same curves differed fundamentally for different types of foods. Curves a, b, and c described the meat curve, d the bread curve, and e and f the milk curve (Figure 1). In sharp contrast to Khizhin, Lobasov concluded that “to each sort of food there is a special character of these changes in the rapidity of secretion of the juice and its qualities.” These patterns—the very existence of which Khizhin (and Pavlov) had denied—were now, for Lobasov (and Pavlov), “repeated with such constancy and such pravil’nost’ that we have the right to distinguish between milk, meat, and bread secretion—and milk, bread, and meat juices.”29
These newly defined curves reflected the development of the factory metaphor and the concomitant “hardening” of interpretations as Khizhin’s often-tentative hypotheses became “quite definite” conclusions. They also reflected the related attempt to incorporate the psyche within the digestive machine. The psyche remained capricious from experiment to experiment, but it had now also acquired a determinist moment: a standard, presumably pravil’nyi psychic response had been identified for each food. 30
For Lobasov, the characteristic secretory curve for each food was the composite of a standard psychic secretion and standard nervous-chemical response to each. Meat produced a strong psychic secretion and was rich in chemical exciters, so the meat curve rose rapidly and declined gradually (as “appetite juice” gave way to nervous-chemical secretion). Bread elicited a strong psychic secretion and was weak in chemical exciters—so the bread curve rose sharply in the first hour and declined rapidly thereafter. Milk elicited a weak psychic secretion and was rich in chemical exciters, so the milk curve rose slowly, reaching its acme only when nervous-chemical secretion kicked in.31
The difference between Khizhin’s (and Pavlov’s) treatment of the psyche in 1894 and Lobasov’s (and Pavlov’s) analysis of 1896 bears emphasis. For Khizhin, Druzhok’s initial response to any food might be of either a “psychic type” or a “usual type.” For Lobasov, each food elicited a standard amount of appetite juice. Appetite is no longer entirely capricious. It is reliably present or absent in the secretory response to specific foods, dependent on the vagus nerves, and subject to physiological mechanisms. So, for example, comparing Druzhok’s secretory responses to whole and nonfat milk, Lobasov identified a determinist mechanism in the glandular consequences of the psyche: fat, he concluded, inhibits the vagus nerves (through which the psyche excites the gastric glands) and so inhibits psychic secretion itself.
Yet the psyche has not been completely subsumed within determinist physiology. It continues to express the character, mood, and food tastes of the dog, and therefore manifests itself differently from trial to trial. Nor do the experimenters explicitly address the essential nature of the psyche. They approach it not as a philosophical question, but rather as an operational, experimental one. The nature of the mind, emotions, and personality per se are black-boxed and employed as an uncontrolled variable. (One dog is greedier than another and so generates more appetite juice; Druzhok is less impressionable or more aggressive one day than another.) They are dealing here with the tension between capriciousness and lawfulness as they seek to acquire pravil’nye results.
In 1896 Pavlov assigned an especially talented coworker, Anton Val’ter, to do for the pancreas what Khizhin and Lobasov had done for the gastric glands. Like Druzhok, Val’ter’s dog Zhuchka was equipped with the latest technology and proved an ideal lab animal. Having recovered her health after implantation of the troublesome pancreatic fistula, Zhuchka quickly regained an excellent appetite and was soon “enjoying” her life. As Val’ter put it: “It is in the steadfastness of this dog, fully stabilized after the operation, that one must find the essential reason for the great pravil’nost’ of the results obtained upon it.”32 Although Val’ter had not yet completed his thesis when Pavlov’s Lectures went to press, his results with Zhuchka already delighted the chief, who used them to display and explicate in his own text the characteristic curves for pancreatic secretion.
The laboratory view of pancreatic secretion was essentially the same as that of gastric secretion, but research on the pancreas confronted additional complications. According to lab doctrine, pancreatic secretion, like gastric secretion, occurred in two stages. In the first stage, the hydrochloric acid in gastric secretion excited the pancreas; in the second, as food moved out of the stomach it excited the specialized nerve endings in the mucous membrane of the duodenum. The specific excitability of these nerves led the pancreas, like the gastric glands, to respond precisely and purposively to specific foods.
Unlike the gastric glands, the pancreas secreted three separate ferments, each of which acted specifically on either proteins, fat, or starch. In a much-praised dissertation of 1893, Vasilii Vasil’ev (and Pavlov) argued that the ferment content of an animal’s pancreatic responses changed over time as the pancreas adapted to its diet.33 In other words, the pancreas not only adapted to the composition of a specific meal during a single feeding, it also underwent a “chronic adaptation” to long-term changes in the animal’s diet. This added another variable to the search for pravil’nye results, since it meant that different dogs, with their different digestive histories, would probably differ in their pancreatic responses to the same meal.
The influence of the psyche on pancreatic secretion was thought to be indirect but omnipresent. There was no experimental evidence for psychic secretion in the pancreas (Val’ter speculated that it existed, but was relatively insignificant). By virtue of the psyche’s important role in gastric secretion, however, it influenced the volume of hydrochloric acid produced in the stomach and therefore, because pancreatic secretion responded sensitively to even small amounts of hydrochloric acid, also played a big role in pancreatic secretion.
Val’ter’s interpretation of experimental trials with Zhuchka led him to the same basic conclusions as had Khizhin’s and, especially, Lobasov’s with Druzhok:
Under identical experimental conditions with the same food, the secretion of pancreatic juice after the dog has consumed milk, bread, or meat is repeated with stereotypical precision. This identity—which is manifest in the course of secretion over time, in its quantity, and in the qualities of the pancreatic juice—testifies to the fact that the secretory apparatus of the pancreatic gland works with great precision and lawfulness.
As with the gastric glands, the precise responses of the pancreas testified to nervous control, to “the specific excitability of the mucous membrane of the gastric-intestinal tract.”34
As with Khizhin and Lobasov, Val’ter’s identification of “stereotypical secretory curves” relied upon his interpretation of the play of uncontrolled variables. These included the water content of the animal’s body, various factors that influenced the amount and strength of the acid in gastric secretion that excited the pancreas, and especially the psyche. He described lucidly the challenges posed by this idiosyncratic actor to the experimenter’s craft and interpretive skills:
In most cases of normal feeding, the first and most powerful exciter of gastric juice is the animal’s appetite—the passionate desire for and enjoyment of food. One must say that this psychic exciter is, by the very nature of things, difficult to subordinate to the control of the experimenter. Even in the purest form of application of this exciter, in “sham-feeding” experiments, the animal often eats with varied interest and therefore produces various quantities of gastric juice; the experimenter’s craft in such cases consists in various tactics, for example, finding a suitable tempo of feeding to hold the animal’s interest at a specific level (usually to arouse it ad maximum). The shortlivedness of normal feeding greatly limits the use of such tactics; here it often occurs that, despite the precise observance of all conditions of the experiment, the animal eats the very same food on various days with varying degrees of enthusiasm. When the dog eats with abandon, there is secreted, at least in the first hours of digestion, a large amount of gastric juice; when the dog eats sluggishly, there is little. Since the work of the pancreas is tightly linked, by virtue of its acidic exciter, to the work of the stomach, the quantity of pancreatic juice also fluctuates in the same manner. This is so aside from the possibility that the psychic moment has a direct effect upon the pancreatic gland.35
Val’ter conceded that experimental results varied widely but, with the chief’s help, deployed the many uncontrolled variables he discusses here to find within them the contours of a precise and purposive mechanism. Zhuchka’s secretory curves thus accompanied Druzhok’s in Pavlov’s synthetic work.
Only afterwards did Pavlov assign two coworkers to verify Val’ter’s findings on other dogs. The disappointingly variant results were attributed in each case to the physiological particularities and personality quirks of these animals. In a passage doubtless coauthored by the chief, one coworker explained that although dogs lacking Zhuchka’s fortunate qualities could not express the same “pravil’nost’ in the work of the gland,” Val’ter’s typology of pancreatic secretions “preserves its correctness even today.”36
* * *
By 1897, the remarkable Druzhok had labored long and well, leaving a considerable legacy to the lab. As the first long-lived survivor of the isolated-sac operation, he had afforded Pavlov and his coworkers their closest look at the normal operation of the gastric glands. Sustaining the subsequent implantation of a gastric fistula and esophagotomy, he had permitted the lab to grapple with the puzzling interaction of psychic and nervous-chemical mechanisms. He had remained throughout “understanding and compliant,” producing results—including so-called “characteristic curves” that reflected, at least in part, his own physical and psychological profile; results that the chief would soon immortalize in a grand synthetic work.
His labors, however, were not over. Just as his survival of the isolated-sac operation had revolutionized one line of investigation, so did the illness that finally resulted from his operations and more than three years of experimental work contribute to the launching of another.
By fall 1897, it was clear to Andrei Volkovich, the third and last coworker to experiment on Druzhok, that the lab’s prize animal was unwell. Druzhok’s isolated stomach had deteriorated so severely as to render him useless for experimentation. Furthermore, his gastric glands had begun to function erratically, leading Volkovich to speculate that the abnormal manner in which the dog had been fed since his esophagotomy (through a fistula, without benefit of appetite) had caused his glands to atrophy.37
The chief had assigned Volkovich the task of comparing Druzhok’s secretory responses with those of the lab’s second dog with an isolated sac, a male setter named Sultan. After several months of experiments, however, Sultan, too, became sick. Volkovich initially paid little attention, “since catarrh is frequent among laboratory dogs,” but the problem worsened and radically altered the dog’s secretory responses. Sultan also became “unusually greedy,” devouring food indiscriminately yet losing weight. When blood began to flow from the isolated sac, his illness was diagnosed as an ulcer of the small stomach—and this was confirmed by autopsy after he died of acute peritonitis.38
The lab, Volkovich noted, had “acquired, completely unexpectedly, the opportunity to observe a clinical case in a laboratory setting.” Experiments on the distorted glandular responses of the sick Sultan to various foods led to the conclusion that hypersecretion was “a neurosis affecting the peripheral endings of the reflexive-secretory nerves.” This neurosis was characterized by the disturbance of “the usual relationships between the psychic and reflexive phases of secretion” and so of the “strict purposiveness” of glandular work.39
The various medical problems with Druzhok, Sultan, and other lab dogs inspired the chief to launch a new line of investigation on “the experimental pathology and therapeutics of digestion.” In the context of this research, Pavlov spoke freely about the illnesses of his lab dogs, drawing upon these as a source of authority in discussions of pathology. In an article on “Laboratory Observations of Pathological Reflexes from the Abdominal Cavity” (1898), he confided that “almost all” the dogs with an isolated sac—by this time, they numbered at least ten—tended to lie on their back with their feet up, indicating they experienced “unpleasant or painful sensations” when in their normal posture. That comment flatly contradicts his assertion in Lectures that the operation did not result in “any sensory unpleasantness,” and also reveals information relevant to judgments about the dogs’ normalcy that was consistently absent from prior lab publications. (The dramatically abnormal image of a dog lying on its back may also explain why the lab did not, in these years, photograph animals with an isolated sac.) Pavlov noted that several dogs with an isolated sac could eat only small portions, and in 1902 intervened confidently in a clinical discussion of suppurating inflammation of the stomach, noting that “not one of the dogs upon which we have operated...has failed to produce similar phenomena.” We cannot know which of these pathologies afflicted Druzhok.40
Volkovich’s and Pavlov’s diagnoses of Druzhok’s illness provide the last bit of information on the dog’s biography. Perhaps the atrophy of his gastric glands did not culminate in a fatal illness; in any case, this veteran experimental animal was incapable of living outside the lab without special care: he was unable to eat normally because of the esophagotomy and leaked gastric juice from the isolated sac for hours after being fed through a fistula. Had special provision been made for him, this would almost certainly have been mentioned by Pavlov’s collaborators as an example of his kindness to experimental animals—especially in view of antivivisectionists’ attacks on him.
The name Druzhok appears in several of the lab’s doctoral theses, but vanishes in the more processed articles of coworkers and in the chief’s synthetic works. The great value of this stellar experimental dog, after all, resided in the generation of knowledge not just about one animal’s psyche and glandular responses, but about the digestive factory in all higher animals. When he appeared again, it was as the anonymous personality behind abstract secretory curves in the chief’s masterwork.