PART I The Nature of the Opiate Habit

CHAPTER 1 METHOD AND PROBLEM

The present study is directly concerned only with addiction to the opiate type drugs and their synthetic equivalents. A rational, general theoretical account of the nature of the experiences which generate the addict's characteristic craving for drugs is proposed and systematically elaborated. While this theoretical position has obvious implications for other addictions which resemble opiate addiction in that they also involve drugs which produce physical dependence and withdrawal distress, an extension of the theory to these other forms of addiction, such as alcoholism, is not attempted here. This is a matter which requires specific empirical investigation of a comparative nature.

The central theoretical problem of this investigation is posed by the fact that some persons who experience the effects of opiate type drugs and use them for a period sufficient to establish physical dependence do not become addicts while others under what appear to be the same conditions do become addicted. The attempt to account for this differential reaction requires a specification of the circumstances under which physical dependence results in addiction and in the absence of which it does not. It also requires a careful consideration of the meaning of "addiction" spelled out in terms of behavior and attitudes characteristic of opiate addicts everywhere. The theory that is developed is a general one; its applicability is not limited to American addicts, to lower-class users, to twentieth-century addiction, to any restricted segment of the problem, or to any specific historical period. Consequently, the focus of theoretical attention must be on those aspects of addiction which may reasonably be regarded as basic or essential in the sense that they are invariably manifested by all types of addicts regardless of place, time, method of use, social class, and other similar variable circumstances.

The second part of the study consists of a brief statement of a view of current public policy concerning addiction in the United States, with proposals for reform which, it is believed, would substantially reduce the evils now associated with addiction and the large illicit traffic in drugs. Since the original publication of this book in IL947 there has been a Widespread public realization that the American program of dealing with addicts is both cruel and ineffective and that it compares unfavorably in these respects with the programs of most European countries in which opiate addiction is handled primarily as a medical rather than as a police problem.

It should be kept in mind that when I use such terms as "narcotics," "drugs," "drug addict," and "drug users," I will be referring to drugs that are commonly classified as opiates and their synthetic equivalents and to persons addicted to drugs in this class. Common and well-known examples are opium, heroin, morphine, methadone, and demerol. Marihuana and cocaine do not belong to this category, and only incidental reference will be made to them. While alcohol is addicting in approximately the same sense that heroin is, it will also be referred to only incidentally. The fact that marihuana, cocaine, and heroin and other opiate type drugs are covered in the same anti-narcotics legislation is a fertile source of confused thinking because it obscures the facts that the use of marihuana is totally unlike heroin or morphine addiction and that alcoholism, which is not covered by the legislation and is not popularly thought of as a form of narcotic addiction, actually has very much in common with opiate addiction.

The theory that is proposed in this book, in its earliest stages, was derived from observing addicts and conversing with them. The later development of the theoretical framework of the study was also significantly influenced by data available in the extensive literature. Approximately fifty addicts were interviewed over a fairly extended period of time sufficient to establish an informal, friendly relationship of mutual trust. In the case of twelve others there was at least one interview but the relationship was brief or occurred under circumstances that made it impossible to place much confidence in what was said or to obtain a full and consistent account of the person's addiction history. Since my contacts were usually with addicts who were using drugs it was unavoidable that some of them disappeared before I could learn much from them. On the other hand, there were some with whom I established relationships that lasted for several years covering periods of use, of temporary voluntary abstention, and of incarceration.

My principal assistance in becoming acquainted with new subjects came from "Broadway" Jones, officially known by the alias "Chic Conwell," which Sutherland gave to him in The Professional Thief.(1) Mr. Jones himself bad bad a long history of addiction that began in his late teens. He was an invaluable -source of information. He also read and criticized my manuscript. Indeed, he initially suggested that I study drug addiction and offered his cooperation in the project. I remained in communication with him until his death, some years after that of Dr. Sutherland. Mr. Jones was not using drugs during most of the approximately twenty years that I knew him but did relapse some years before his death. On the advice of Sutherland he voluntarily committed himself to Lexington for withdrawal.

When I first began to talk to addicts with the idea of making a systematic analysis of the habit I became fascinated by the theoretical problems it posed. I dreamed of basing my analysis on interviews with hundreds of addicts, perhaps even as many as a thousand. For this purpose, through the good offices of Professor Sutherland I obtained permission from the Federal Bureau of Prisons to interview addicts at the Lexington Public Health Service Hospital and the annex of the Federal Penitentiary at Leavenworth, which at that time specialized in handling drug offenders. However, the Public Health Service, which also exercised authority in these two establishments, refused to give me this permission on the grounds that what I proposed to do was already being done by their personnel. I was consequently compelled to rely on the costly, slow, and arduous techniques involved in picking up subjects from the Chicago streets, financing my research almost entirely from my own pocket.

Many but not all of my subjects were introduced to me by Broadway Jones, who briefed both them and me in advance. Most of them were experienced in criminal activity such as drug peddling, some form of theft, or the confidence game. Mr. Jones sometimes warned me that my new subject could not be trusted and that I should keep one hand on my billfold and the other on my watch when I was with him. In other instances the subject was characterized as an "honest thief" who could be trusted implicitly to steal only from strangers and whom I could safely take into my home. A substantial number of my subjects did in fact visit with me, my wife, and our daughter in our apartment. None of these ever pilfered anything; on the contrary, a number of them offered to provide us with stolen goods.

I ordinarily told each new subject that I was interested in studying the "psychology" of the habit and trying to understand it. This invariably satisfied them, and they asked no probing questions, perhaps because they felt that they knew a great deal more about the psychology of addiction than I ever would. It was thus unnecessary to explain the theoretical development of the study to them, and, indeed, they were not interested in it.

To encourage addicts to remain in contact with me I usually rewarded them by buying them a meal in a cheap restaurant or bar or by giving them fifty cents or a dollar after an interview. My conversations with them were held in bars and restaurants as a rule. Most of my subjects were very cooperative, in part perhaps because addicts talk and think about their addiction a great deal in any case, and also, in many instances, because they felt that an objective, non-moralistic study of addiction was needed and might benefit them by leading to a more rational public policy. There is a widespread feeling among addicts that addiction is not understood, that it is misrepresented in the mass media, and that it is dealt with in an inhumane and needlessly severe manner. Non addicts who have dealings with known addicts generally deal with them in a condescending or authoritarian manner and often moralize with them. I consciously avoided all of this, even at times when I was repelled or shocked. I felt no condescension because I knew that my subjects had information which might be useful to me and that I was thus dependent on them. Finally, I was in no position to be authoritarian because my subjects were free to break off the relationship whenever they wished to.

The literature of drug addiction was at first not consulted for fear that the opinions expressed would introduce an initial and perhaps decisive bias into the investigation. The first temporary hypothesis was formulated exclusively on the basis of observation of and conversations with addicts, and it was only after the study had crystallized around a few central problems and theories that the literature was intensively examined. It was examined particularly with the view of uncovering negative evidence which would force revision of the theory and also of exploring as exhaustively as possible all relevant aspects of the subject and all implications of the theory.

The first tentative and obviously inadequate hypothesis formulated was that individuals who do not know what drug they are receiving do not become addicted and, on the positive side, that they become addicted when they know what they are getting and have taken it long enough to experience withdrawal distress when they stop. This hypothesis was destroyed almost at once by negative evidence. One of the first addicts to be interviewed, a doctor, had once received morphine for several weeks; be was fully aware of the fact, but be did not become addicted at that time. The difficulty presented by this case remained unresolved until my attention was attracted to a rather casual and incidental comment made by Dr. Albrecht Erlenmeyer in an article concerned mainly with the physiological effects of morphine. Speaking of withdrawal distress, Erlenmeyer said, "In such moments the craving for morphine is born and rapidly becomes insatiable, because the patient has learned during the period of habituation, when abstinence symptoms always set in after the effect of the last morphine dose has passed off, that those terrible symptoms are banished as if by magic by a sufficiently large dose of morphine ."(2) In the light of this statement, the second hypothesis of the investigation was that persons become addicts when they recognize or perceive the significance of withdrawal distress which they are experiencing, and that if they do not recognize withdrawal distress they do not become addicts regardless of other considerations.

This formulation proved to be much more significant and useful than the first, but like the first it did not stand the test of evidence and had to be revised when cases were found in which individuals who bad experienced and understood withdrawal distress, though not in its severest form, did not use the drug to alleviate the distress and never became addicts. The final revision of the hypothesis involved a shift in emphasis from the individual's recognition of withdrawal distress to his use of the drug to alleviate the distress after this insight has occurred. This hypothesis was found to be superior to the others. It had the advantage of attributing the origin of addiction, not to a single event, but to a series of events, thus implying that addiction is established in a learning process extending over a period of time.

If one thinks of the central problem of addiction as being that of isolating and describing the nature of the experience from which the fatal craving or "book" is derived, the hypothesis stated above suggests that the critical experience in the fixation process is not the positive euphoria produced by the drug but rather the relief of the pain that invariably appears when a physically dependent person stops using the drug. This experience becomes critical, however, only when an additional indispensable element in the situation is taken into account, namely, a cognitive one. The individual not only must experience relief of withdrawal distress but must understand or conceptualize this experience in a particular way. He must realize that his distress is produced by the interruption of prior regular use of the drug.

It is evident that when a person becomes addicted to heroin or morphine the entire pattern of his social behavior is commonly altered to a pervasive degree and that his orientation toward the drug is radically altered. These drastic changes in behavior and attitude are not of the kind that could reasonably be expected to occur in an instant. They are learned, according to the theory, in a gradual but rather rapid way as the withdrawal distress recurs and must repeatedly be banished by further use of the drug. Even though the beginner may like the effects of morphine or heroin, the prospects of becoming an addict are traumatic if they are actually fully understood. In this sense, the addict does not ordinarily become such voluntarily but is rather trapped "against his will" by the book of withdrawal.

If one disregards the cognitive aspect, one can say of this hypothesis that it is a theory emphasizing negative rather than positive reinforcement as the basis of addiction. In psychological writings the term 11 negative reinforcement" is used to refer to a situation in which a given action is rewarded by the elimination of something unpleasant-an adverse stimulus. Thus, if a child is locked in his room until be masters his school assignment, his reward for doing it will be that he will be freed from confinement. This is negative reinforcement. If, on the other band, the child is promised a much desired trip to the zoo when be completes his homework, this is called positive reinforcement.

It will be pointed out in a later chapter that psychologists have explored the theory advanced here, minus the cognitive feature, in ingenious experimental investigations of the effects of opiates on lower animals. Some of this work strongly suggests that insofar as lower animals are capable of matching the behavior of human subjects in becoming attached to drugs, their attachment also depends upon the negative reinforcement obtained from the relief of withdrawal rather than on positive reinforcement stemming from pleasurable effects of the drugs. Investigations of this type which necessarily ignore the differences between human beings and animals occasioned by the former's immense cognitive superiority leave open the question of the extent to which the responses of lower animals to opiates can be made to parallel those of human subjects. In a subsequent chapter attention will be given to the matter of determining whether it is justifiable or meaningful to assert, as some investigators do, that addiction can be induced in rats, monkeys, and other animals.

The hypothesis of this study was tested under the following assumptions: ( 1) the verification of a theory consists not in piling up selected instances which confirm it but in looking for evidence which contradicts clear logical implications which may be deduced from it; (2) a valid theory of addiction must account for the basic or essential aspects of addiction by indicating that they form a system or pattern which is logically implied or predicted by the theory. All of the evidence obtained from the literature and from conversations with all subjects was considered from this point of view. None of it seems to me to contradict the theory. On the contrary, the theory seems to make sense of a number of aspects of addiction which have usually been regarded as paradoxical or puzzling from other points of view. It also suggests a simple unitary explanation of a central theoretical problem, that of accounting for the fact that physical dependence on opiates is sometimes followed by addiction and sometimes is not.

The fact that the hypothesis of the study was revised a number of times in the course of the investigation suggests that further evidence or the extension of the theory to other forms of addiction may necessitate further reformulations. This is a probability rather than a mere possibility and seems to me to constitute an advantage rather than a weakness. It is characteristic of all genuine scientific systems that they evolve as they are confronted with new evidence that does not fit the old theories. The prime virtue of a general theory is that it stimulates the search for negative evidence and challenges its critics to construct a better one.

It may be asked whether the search for negative cases was properly conducted and if the observer has not neglected evidence of a contradictory character. To this, of course, there is no final answer. It is probable that somewhere in the course of any study unconscious distortion takes place. Concerning the central hypothesis and the direct lines of evidence, however, certain procedures were followed which may be said to exclude bias. For example, when the theory had been stated in an approximation of its final form it occurred to me that it could be tested in cases where an individual bad bad two separate experiences with morphine or opiates, each of which was sufficiently prolonged to produce withdrawal distress but with addiction following only the second episode. Case 3 in Chapter 4 is an example. It was concluded that if the theory was valid, the person would report that he bad failed to realize the nature of the withdrawal in that experience from which he had escaped without becoming addicted. Thereupon a thorough search was made for cases in which an individual had undergone such an experience with the drug prior to becoming an addict. All cases of this kind which could be found, or of which any record could be located, were taken into account. Any of these cases might have contradicted the final hypothesis, but none did so. The inference or prediction which had been drawn on the basis of the theory was thus fully borne out. This procedure was followed throughout the study wherever possible and, as will be seen, is implicit in the form in which the theory is stated.

As previously mentioned, I scrupulously refrained from informing the addict of any theories which I held or of the exact reasons for my inquiries. Thus when an addict was asked about his experience during gradual withdrawal cures he was not told that an attempt was being made to check the implications of the theory. It was inferred that if addicts feel approximately normal between shots as they maintain they do' it should be possible to deceive them as to whether they are really getting the drug. During the discussion of such gradual withdrawal cures, the information was often volunteered by the addict that be himself bad been deceived at some time into believing that he was getting the drug when he actually was not, or vice versa. This information was usually introduced with the remark that he did not expect to be believed, whereupon be would proceed to give the facts which corroborated the theory and which had been anticipated. In none of these cases, however, did the addict know when be gave the information that it had any significance for a theory of addiction.

It has often been said that testimony of addicts is virtually worthless because of the secrecy surrounding addiction and the tendency of drug users to distort and falsify. I have found that this view is incorrect. The addict lies and distorts not because be has any indiscriminate urge to do so but in order to obtain certain definite and practical results. If it is clear to him that be has nothing to gain by lying and nothing to lose by telling the truth, he is as straightforward and honest as anyone else. Of course, he will rationalize and offer excuses, but in this sense be is no different from non-addicts.

It is true that there are certain areas in which the addict has very strong inhibitions against giving the unvarnished truth, but these inhibitions apply entirely to information which was not relevant to the central problem of the study. A drug user is particularly cautious about information concerning the sources of his supply, and if he is engaged in theft or other illegal activity he will be very reluctant to describe this activity. Since matters of this kind had no bearing on the central problem being investigated no special effort was made to obtain these data, although practically all the addicts who were interviewed eventually volunteered such information. The drug user has no special tendency to falsify concerning the nature of the drug habit. The fact that contact was maintained with some of the cases over a period of several years was a further insurance against deception. On the whole, once friendly contact is made, drug users are probably above average in cooperativeness, because many of them believe that they ought to be studied and better understood. The data which they provided me were found to be quite' consistent with the material that the literature on opiate addiction offers.

This study might be criticized on the grounds that the addicts who were interviewed did not constitute a representative sampling of American drug users. Indeed, no effort whatever was made to secure a sample that was representative in a statistical sense because it was not the purpose of the study to make a statistical description of the variable attributes and characteristics of American addicts. The purpose was rather to try. to find a plausible general theory of addiction, that is, to describe the nature of the experience from which the user's craving for the drug is derived.

The investigation of any subject necessarily proceeds on the basis of assumptions that cannot themselves be established as valid in the research in question. One such assumption which underlies the present study is that the scientist who studies social or other phenomena should, whenever possible, seek to formulate generalizations that apply to all of the instances of the problem with which be is concerned, rather than to most or some of them only. In other words, it is assumed that genuinely scientific causal propositions should be stated as universals. This does not imply a belief in absolute truth as is sometimes erroneously assumed, since it is anticipated that all such generalizations will be accepted provisionally only, as long as no contradictory evidence is available or no better theory is at hand. This assumption implies the belief that it is possible for the social scientist as well as for the natural scientist to state theories in such a form as to suggest crucial instances testing the theory and to permit the search for negative cases. It is assumed, in other words, that the exceptional instance is the growing point of science and that cumulative growth and progressive development of theory is obtained by formulating generalizations in such a way that negative cases force us either to reject or to revise them.(3)

The methodological orientation of this study is deterministic rather than statistical. Applied to the study of narcotics addiction, each of these orientations has its own characteristic goals, methodological assumptions, and analytical procedures, and neither can be judged by the standards of the other. Failure to distinguish between them leads to intellectual confusion.

The assumption made at the outset of this investigation-that the craving for drugs is generated in one identifiable, unitary type of experience-is deterministic in nature. The research problem it poses is that of identifying or discovering and describing such an experience and indicating bow and why the patterns of behavior called addiction follow from it. It is incongruous and illogical to consider such an experience as a "variable" to be handled by' statistical methods. This is because an "experience" is a complex interactional process- involving many elements or variables in a series of happenings or events. Moreover, if one insists on speaking of such a sequential process as a variable, it must be noted that it is not at band at the outset of the study since it has not yet been identified and described. From the deterministic viewpoint the whole trick consists of finding or identifying it. To suggest that this identification be made by statistical methods presupposes the knowledge that is being sought.

It is more appropriate to consider the statistical approach to drug addiction within its own frame of reference as a technique of describing average behavior of aggregates of individuals or of making mass comparisons between groups of addicts and non addicts. Such comparisons and descriptions have dealt with a wide variety of attributes, characteristics, and background factors such as those of age, sex, social status, place of residence, occupation, and personality attributes. These investigations have their own particular kind of merit or significance. They may, for example, illuminate such matters as why addiction rates are higher in one segment of the population than in another; they may indicate that certain identifiable personality traits increase the probability of addiction-, they may suggest that public policies of dealing with addiction have important consequences on the prevalence of addiction and on variations in its incidence. Such studies are often of great practical importance. They may also be of theoretical significance at their own level-that of broad "social system" or societal analysis. They may suggest ideas to the determinist, and their findings must be taken into account, but they do not have a direct bearing on the determinist's search for the causal process or processes that always and everywhere generate addiction. The difference between these two types of enterprise is analogous to that between explaining malaria and explaining malaria rates, or crime and crime rates. From a deterministic standpoint, conclusions derived from statistical comparisons of addicts with non-addicts which assert as a "theory" that a given attribute is significantly more frequent among drug users than in the general population are simply not theories at all.

The attempt to identify causal relationships and the assumption that such relationships can be found dictate that all relevant and available evidence, including the results of statistical studies, be taken into consideration. In particular, this methodological stance requires intensive, exhaustive probing of individual cases and comparison of certain crucial types of cases. Thus, when it was noted that some non addicts receive drugs regularly for long periods of time in hospitals without becoming addicts, I was compelled to compare them with addicts in order to isolate and describe the causal processes which were present in the cases of addiction and absent among the non-addicted hospital patients. No tabulations were necessary in making this comparison because it was assumed that the essential causal process of addiction would be found in all cases of addiction and that it would not be found in any case of non-addiction.

The addicts who were interviewed contributed very unequally to the final theoretical formulation. As already indicated, some were not seen often enough. Others were inarticulate or lacked the necessary intelligence or interest to provide coherent accounts of their initial experiences with drugs. Many who were interviewed in the later stages of the investigation provided data of relatively little theoretical import because the information they gave followed a pattern made familiar by earlier cases without adding new elements or posing new problems. Only a relatively small proportion of the addicts interviewed, consisting mainly of intelligent, articulate, experienced, and self-observant users, made important contributions to the formulation of the theory. Some of these crucial cases forced the abandonment of provisional hypotheses that had been entertained up to the time they were encountered. Others seemed to bring out in a striking and obvious way the nature of the process in which addiction is established. This was because they were departures from the standard and familiar pattern, presenting combinations of circumstances and conditions of the type which one would wish to have in an experimental test of the theory. Aft& familiarity with the general characteristics of the addict's behavior bad been acquired from the initial conversations with users, the subsequent progress of the study seemed to depend upon the analysis of a series of crucial cases which led to successive revisions of the guiding ideas of the study and to broader perceptions of the logical implications and ramifications of these ideas. As the analysis progressed the various aspects of addiction behavior, which had at first seemed to be isolated, discrete bits of information or even paradoxical in nature, fell into place to form integral parts of what eventually seemed to me to be a consistent and logical whole.

The literature of addiction is highly repetitive, sometimes unreliable, and often based on misinformation. In the later stages of the study, however, it probably became my most important source of data. The interviews with addicts and observations of them provided standards for judging the literature and enabled me to understand the sources of error and distortion. The evidence available in books and articles served the vital function of providing a broad, world-wide perspective, limited by neither space nor time. Without such a perspective it would have been easy to fall into the error of taking the part for the whole, that is, of assuming that a particular manifestation or form of addiction limited to a particular time or place was the prototype of all addiction. Thus, no general theory of opiate addiction can be based on effects which are produced only when heroin is injected intravenously, as it usually is at present in this country, since opiate addiction existed many thousands of years before heroin or the intravenous method of injection were known.

The strategy of the research was one of focusing on a strictly limited and specific problem, analyzing it as exhaustively as possible, and leaving out of consideration a great many other matters that might be of interest and significance in connection with other types of theories or interests.

This study, it will be observed, is not a study in motivation and therefore does not propose an explanation of addiction in terms of the motives or purposes that people have for trying drugs. It began as a search for the experience or process in which addiction is invariably generated, without any prejudgments concerning the motivation question. As it progressed it seemed evident that no general theory of addiction in terms of motives was possible. This was because: (1) hospital patients sometimes become addicted when doctors prescribe drugs for them without the patient ever voluntarily administering the drug to himself; (2) motives or verbal rationalizations of drug users vary widely from person to person, from culture to culture, and from time to time, and they are often contradictory and inconsistent; and (3) while initial use of the drug is often but not always voluntary, non addicts are rarely motivated to become addicts, but are rather trapped by the drug, often unwittingly or against their wills. As will be indicated later, the initial effects of drugs upon the beginner are not at all comparable to the effects of the same drug upon the habitual user, and, by the same token, explaining the first trial is a separate and distinct problem from that of accounting for addiction.

While motives are not considered as causes of addiction in this inquiry they are not altogether irrelevant, since the study might be characterized as being concerned with the origins of the addict's drive to obtain and use drugs. Once established, this drive becomes a dominant feature of the user's motivational life and tends to extend its influence to virtually all of his behavior. The intricacies of this revolutionary change in personality are an appropriate subject for psychiatric exploration and beyond the limits of this investigation.

The conventional view of addiction is that it is an escape mechanism for defective persons, variously characterized as "inadequate," "inferior," "frustrated," or "psychopathic." It is assumed that persons become addicts for the "Purpose" or "motive" of escaping from or alleviating their psychic troubles. This confusion of "cause" and "motive" leads to difficulty because of two stubborn facts which cannot be ignored or explained away: (1) some persons become addicts as a consequence of medical practice under conditions which preclude the influence of their motives on any part of the process of becoming addicted, and (2) a substantial percentage of addicts are admittedly "normal" prior to addiction; that is, no evidence of defects, inferiority feelings, inadequacy, etc., can be found. The proponents of this conventional view satisfy themselves in this situation by asserting that most addicts are abnormal prior to addiction and ignore or write off in advance a considerable body of exceptional cases. No satisfactory explanation is offered for the fact that "normal" individuals become addicts. It is even admitted sometimes that the generalization probably does not apply even to most addicts in some parts of the world.

Since exactly the same explanation is applied to many other forms of undesirable behavior, it has no particular or specific explanatory power with regard to any of them. The defective individual is sometimes pictured as one who casts about, reviewing alternative forms of deviant behavior and choosing one more or less at random or on the basis of availability. A theory of this sort has the attraction of seeming to explain much with little investigative effort. Applied to drug addiction it has the added charm of making it superfluous to bother about examining the actual mechanisms involved.

The logical inadequacies of this type of approach are obvious. If the validity of a theory is defined in terms of the empirical evidence that might demonstrate its falsity but does not, this theory is unverifiable, since the exceptions that negate it are admitted in advance. If one is concerned only with some, many, or most addicts, there are many variables in addition to personality factors that are positively associated with addiction rates. These variables differ from one country to another; they differ from one group to another and from one place to another; and they change with time even in the same country. It is therefore possible to formulate a series of theories, each of them applicable to some addicts in some place at some time, and none of them applicable to all addicts anywhere. Under such circumstances DO grounds exist for saying any such theory is wrong, that one is right and the others wrong, or that one is any better than another. Acceptance of one above the others becomes a matter of personal taste or professional prejudice.

The point of view under consideration is based on preoccupation with variability and with statistical techniques of dealing with it. It involves a fundamentally skeptical or negative attitude toward the possibility of discovering ordered patterns in the behavior of the individual. Human behavior is said to be too complex, too dynamic, too indeterminate, and to vary too much from culture to culture and from person to person to be dealt with in any other way than in the aggregate or on the average. Kurt Lewin characterizes this viewpoint as Aristotelian, contrasting it with the Galilean position, which he views as characteristic of modern science:

The conviction that it is impossible wholly to comprehend the individual case as such implies ... a certain laxity of research: it is satisfied with setting forth mere regularities. The demands of psychology upon the stringency of its propositions go no further than to require a validity "in general" or "on the average" or "as a rule." The "complexity" and "transitory nature" of life processes make it unreasonable, it is said, to require complete, exceptionless validity. According to the old saw that "the exception proves the rule," psychology does not regard exceptions as counter-arguments so long as their frequency is not too great.
The attitude of psychology toward the concept of lawfulness also shows clearly and strikingly the Aristotelian character of its mode of thought. It is founded on a very meager confidence in the lawfulness of psychological events and has for the investigator the added charm of not requiring too high a standard of validity in his propositions or in his proofs of them. . . .
Methodologically also the thesis of the exceptionless validity of psychological laws has a far-reaching significance. It leads to an extraordinary increase in the demands made upon proof. It is no longer possible to take exceptions lightly. They do not in any way "prove the rule," but on the contrary are completely valid disproofs, even though they are rare; indeed, so long as one single exception is demonstrable. The thesis of general validity permits of no exceptions in the entire realm of the psychic, whether of child or adult, whether in normal or pathological psychology.(4)

Writing on the nature of scientific systems, Braithwaite makes the following comment:

The one thing upon which everyone agrees is that it (a scientific law] always includes a generalization, i.e., a proposition asserting a universal connexion between properties. It always includes a proposition stating that every event or thing of a certain sort either has a certain property or stands in certain relations to other events or things, having certain properties.(5)

The significance of theories formulated in universal form, as emphasized by Lewin, Braithwaite, and many others who have concerned themselves with the logic of science, depends less on their truth than on the functions they perform in guiding the search for evidence, providing guidelines in the analytical process, and providing means for testing theories. A proposition or theory which affirms an invariable or universal connection between events, unlike one which affirms the connection in only some of the instances, lends itself much more readily than the latter to deductive elaboration and hence to falsification by negative evidence. Because it claims much, it is easily disproved if it is in fact false.

Once a beginning has been made in the construction of theories of this type there is a built-in guarantee of progressive evolution, since evidence not consistent with the existing. theoretical structure forces changes in the structure, improving and broadening it.The progressive refinement of theory that is brought about by thenecessity of taking negative instances seriously regardless of theirfrequency also makes for a progressively closer articulation oftheory with the empirical evidence generated by research.

Another important function of theory couched in the universal form is that it directs research efforts to the problematic areas in which its validity may be in doubt. Because the verification of a scientific theory consists primarily in the failure to prove that it is wrong, it is most effectively tested by the close examination of precisely those areas in which it seems the weakest or of the instances which appear to contradict it. This is why, in the advanced sciences, investigators do not simply go on replicating endlessly the experiments on which the established theories are based. Instead, new logical implications of the currently accepted scheme are traced into new areas where They are confronted by new evidence and tested by new techniques.

Writing of a method which he called " analytic induction," Znaniecki said:

The only reason for the existence of this method (analytic induction) is, as we have seen, the impossibility of arriving by enumerative induction at judgments of the type "All S are P," or rather "if p then 2." Whenever such judgments can be formulated, the use of the statistical method is precluded. Now everybody knows that in physical and biological sciences there are innumerable judgments bearing upon empirical reality which have this form. Every botanist or zoologist in describing a species means to characterize all the living beings of this species; every physicist or chemist in formulating a law claims that the law is applicable to all the processes of a certain kind. Does this mean that the reality with which the botanist or the physicist is dealing is so uniform empirically that no exceptions, no deviations from the type can ever be observed.? Of course not; the botanist or the physicist well knows that cases may be discovered which will contradict his generalization, But he is not afraid of them. He is ready to grant any exception as raising a problem and thus stimulating new research. The research may enlarge and confirm his theory by helping to discover a new species or a new law definitely connected with his previous generalization-which will mean that the exception was only apparent. Or further research may invalidate his former generalization and force him to reach wider and deeper in creating a new and more efficient theory.(6)

It should be observed that Znaniecki neither invented this method nor was he by any means the first to describe it. The distinction between enumerative and analytical induction is derived from one of the oldest and thorniest problems in the logic of induction and is made by numerous writers, including for example, John A Keynes, who grappled with it using the terms "pure induction" and "induction by analogy." The distinction arises from the fact that some of the relatively secure conclusions of science have been derived from the examination of very few instances, as for example, from a few repetitions of a crucial experiment or even from one such experiment. On the other hand, highly questionable, unreliable, and even false propositions have been based on the examination of very many instances. "Pure' or "enumerative" induction refers to the type of analysis in which the dependability of one's inferences depends solely on the number of instances examined. "Induction by analogy" or "analytic induction" refers to another mode in which the validity of conclusions depends on something other than the sheer number of examined instances, such as their variety, the extent of one's knowledge of each of them, and the circumstances presented by each. The crucial experiment of the classical form may be taken as a prototype of the latter mode.

Such terms as "cause" and "causal process" have been used as a matter of convenience in this discussion and will be used repeatedly in subsequent pages to refer to the special kind of relationships that are affirmed by universal theories. The substance of the methodological position that has been briefly sketched here could have been described without the use of these terms, just as there are many inquiries made within the deterministic framework which never make use of them.

On the other side of the fence, some social science statisticians have recently made considerable use of the concept "cause," giving it a very different meaning from that which it has traditionally had or that it has here. Others reject the concept altogether and urge that the latest developments in sub-atomic physics have totally discredited it along with the entire logic and philosophy of nonstatistical determinism. Since this drastic conclusion is currently under debate and is rejected by many eminent natural scientists and logicians, it is premature for social scientists to accept it as given. Considering the vast differences between the theoretical sophistication of modem physics and modern sociology, the rejection of causality and determinism by the latter looks more like an attempt to dodge problems than to solve them.

It is possible that the lesson which the social scientist might learn from these statistical theories of modern physics is the more modest one that strict causality and determinism are inappropriate concepts at that level of analysis in which individual instances are lost sight of and are merged in the description of aggregates. Since individual human beings, unlike electrons, are accessible to direct observation it is unnecessary to study them only in the aggregate.

1. E. H. Sutberland,The Professional Thief (Chicago: University of Chicago Press, 1937).

2. As quoted by Charles E. Terry and Mildred Pellens,The Opium Problem (New York: Committee on Drug Addictions and the Bureau of Social Hygiene, 1928), p. 602.

3. George H. Mead, "Scientific Method and Individual Thinker," in John Dewey (Ed.), Creative Intelligence (New York: Henry Holt, 1917); A. D. Ritchie, Scientific Method. An Inquiry into the Character and Validity of Natural Laws (New York: Harcourt, Brace, 1.923), PP. 53-83.

4. Kurt Lewin,A Dynamic Theory of Personality (New York: McGraw-,?Tm, 1935), pp. i8-ig, 24.

5. R. B. Braithwaite,Scientific Explanation: A Study of the Function of Theory, Probability and Law in Science (New York: Harper Torchbooks, ig6o), p.9.

6. Florian Znaniecki, The Method of Sociology (New York: Rinehart, 1934), PP. 232-33.