Chapter 4 III Criticism of Anti-Naturalistic Doctrines

19.practical significance of criticism Whether scientific inquiry is really motivated by a desire to know, that is, purely theoretical or mere curiosity, or whether we should understand science as a tool for solving practical problems in the human struggle for existence is a question that does not have to be discussed here. solved problem.It can be argued that those who defend the right to "pure" or "fundamental" research should have every support against the narrow and unfortunate fashionable view that scientific research is only valid if it is a legitimate investment.But even the somewhat extreme view (which I am inclined to) that science is of the utmost importance because it is one of the greatest spiritual adventures known to man can at the same time acknowledge practical issues and for the sake of The importance of practical testing for the progress of science, whether applied or pure, for practice is invaluable to scientific thinking, both as spurs and as bridle.We need not adopt pragmatism in order to appreciate Kant's words: "Allow us to have every curious whim, and let our zeal for inquiry be restrained in no way except that which is within our reach, that is what keeps us from being pedantic." It is a spiritual pursuit. The ability to discern, among the innumerable problems that arise, those problems whose solutions will be of vital importance to mankind is wisdom".

This idea clearly applies to the biological sciences and even the social sciences.Pasteur's reforms in the biological sciences were inspired by very practical problems, some of them industrial and agricultural.The urgency of research in today's society even exceeds that of cancer research.As Professor HayeK said: "Economic analysis has never been the product of detached intellectual curiosity to explore why social phenomena are the way they are, but the result of a strong desire to transform an extremely unfair world. , some other social sciences have not yet taken this view, and their fruitlessness shows how urgently their thinking needs practical testing.

When we look deeply at the method of scientific research, especially the general or theoretical social science method we are dealing with here, it is obvious that the stimulation of practical problems is also needed. Arguments about method that do not arise from practical problems are almost useless scrutiny, and this situation makes methodology look down on practical researchers.It should be recognized, however, that more practical methodological arguments are not only useful but necessary.In the development and improvement of methods, just as science itself develops and improves, we can only learn by trial and error, and we need criticism from others to discover our mistakes; since the introduction of new methods implies fundamental and revolutionary change, so this critique is all the more important.The introduction of mathematical methods into economics, the introduction of so-called "subjective" or "psychological" methods into value theory, etc., are examples.A recent example is the combination of value doctrine with statistical methods (needs analysis).This new revolution in method is in part the result of a long and prolific critical debate; defenders of method studies will certainly take heart from this fact.

A practical approach to the study of social science and its methods is advocated by many followers of historicism, who hope that they can use historic methods to transform social science into a powerful tool in the hands of politicians.It is this recognition of the practical tasks of the social sciences that provides a common ground for discussion between the historicists and some of their opponents; on this common ground I am prepared to take my position and thereby criticize the historically determined The theory is a clumsy method that does not produce the results it promises. 20.Technical Methods in Sociology

Although here my subject is the methodological doctrine of historicism which I do not agree with.Not those methods which I consider to have been successful and wish to have a further and more conscious development, but here first briefly touch on those which have succeeded, in order to indicate to the reader my own tendencies, and to clarify the grounds on which my criticism is based. point of view, there are benefits.For convenience, I will refer to these methods as "piecemeal technology". The term "social technology" (and "social engineering", introduced in the next section) can be suspicious and possibly offensive to those who, when they hear it, think of collectivist planners or The social blueprint of the "technocrats".I am aware of this danger, so I add the word "incremental" to remove unpleasant associations and express my belief that "incremental repair" (as it is sometimes called) combined with critical analysis is The social sciences as well as the natural sciences are the main means of achieving practical results.The social sciences have made enormous advances by criticizing and suggesting improvements in society, or, more precisely, by trying to find out whether certain economic or political actions will produce the desired or desired results.This approach could indeed be called the classical approach, and this is what I have in mind when I attribute the technical approach to the social sciences or to "incremental engineering."

In the social sciences, technical problems can be of a "private" or "public" nature.For example, research on the effects of business management techniques or improved labor conditions on production falls under the former category.The effect on the average of incomes of prison reform or universal health insurance or of the use of the courts to stabilize prices, or of new import duties, etc., is of the latter; The question of whether central "planning" in the sense that the state manages production is compatible with effective administrative democratic oversight; or how to export democracy to the Middle East, also falls into this category.

The emphasis on practical technical approaches does not mean that theoretical questions raised through the analysis of practical problems should be excluded.On the contrary, one of my main points is that a technical approach may help to ask a great question that is purely theoretical.But the technical method, besides the basic task of helping us to choose our problems, imposes a precept on our purely theoretical tendencies (especially in the field of sociology itself, which easily lead us into the realm of metaphysics). ); for it obliges us to subject our theories to definite standards, such as those of clarity and practical testability.I might be able to formulate my view on technical methods as follows: Sociology (and perhaps social science in general) should seek not so much "its Newton or its Darwin" as "its Galileo or its Pasteur."

This, and the analogy I mentioned above between the methods of the social sciences and the methods of the natural sciences, is likely to be objected to, as is the case with our choice of words like "social technology" and "social engineering" (although The word "progressive" expresses important constraints).I would therefore best say that I fully understand the importance of the struggle against rigid methodological naturalism or "scientism" (to use Professor Hayek's term).However, it is not clear to me why we cannot make use of this productive analogy, even though we recognize that it has been grossly abused and misinterpreted in some quarters.Besides, we have shown that some of the methods attacked by these rigid naturalists are essentially the methods employed in the natural sciences, and it is difficult to raise a stronger argument against them.

A superficial objection to what we call a technological approach is that it implies an "activist" attitude towards social order (see Section 1), and thus tends to disapprove of our anti-interventionist Or the "passivist" point of view.The view is that if we are dissatisfied with existing socioeconomic conditions, it is because we do not understand how they work and why active intervention can only make things worse.Here I must admit that I certainly do not share this "passivist" view in the slightest, and I also think that a comprehensive anti-interventionist policy is untenable - even on purely logical grounds, since its Advocates have to recommend some kind of political intervention in order to prevent it.However, the technical approach itself is neutral on this issue (as it should be) and is incompatible with anti-interventionism; To make matters worse is to say that certain political actions fail to achieve certain results—not the desired results.One of the most typical tasks of any technology is to figure out what cannot be achieved.

It is worth considering this more closely.As I have shown elsewhere, all laws of nature can be expressed as asserting that something cannot happen; that is, in the proverbial way of saying "You cannot draw water from a bamboo basket".For example: the law of conservation of energy can be stated as: "You cannot make a perpetual motion machine", and the law of thermodynamics can be stated as: "You cannot make a machine with 100% efficiency".This way of elaborating the laws of nature makes its technical meaning apparent, so it can be called the "technical form" of the laws of nature.If we consider anti-interventionism in this sense, we see right away that it could well be expressed in terms of: "You can't have this result," or, "Without this concomitant effect, you would It can’t be that way.” But it shows that anti-interventionism can be called quintessentially technocratic.

Of course, the field of social sciences is much more than that.Rather, the significance of our analysis lies in the fact that it draws our attention to a really fundamental similarity between the natural and social sciences.I think that social science laws or hypotheses exist, and they are similar to natural science laws or hypotheses.Since the existence of these sociological laws or hypotheses (as distinct from so-called "historical laws") has often been doubted, I will now give some examples: "You cannot introduce an agricultural tariff and at the same time reduce the cost of living,"—"In an industrial You cannot organize consumer pressure groups as effectively in society as you organize producer pressure groups.”—You cannot have a centrally planned society while maintaining a price system that has the primary function of competitive prices. ’ — “You can’t have full employment without inflation. ’ Another set of examples can be taken from power politics: ‘You can’t have a political reform without having an effect that is undesirable for the ends that are sought. (Thus, beware of them). -- "You cannot carry out a political reform without intensifying the opposition to a degree roughly proportional to the extent of the reform." (This can also be said to be a technical corollary of "there are always interests connected to reality") -- —"You cannot make a revolution without arousing rebellion." Two more may be added to these examples, which may be called "Plato's Law of Revolution" (from Book VIII) and "Lord Acton's The Laws of Corruption", respectively: You cannot have a successful revolution if the ruling class is not weakened by infighting or defeat. "—"You cannot give a man power over others without inducing him to abuse it"—this temptation increases roughly with the amount of power possessed, and almost no one can resist it ’” Here we do not assume that the strength of the available evidence supports these hypotheses, the formulation of which leaves much room for improvement.They are but examples of the type of statements that progressive techniques are meant to discuss and flesh out. twenty one.Incremental Engineering versus Utopian Engineering Despite the controversial associations that the word "engineering" brings to bear, I use "incremental social engineering" to describe the practical application of incremental technological results.The usefulness of the term arises from the need for a term for all kinds of social activities (private and public) which, in order to achieve a purpose, make conscious use of all available technical knowledge.Incremental social engineering, like natural engineering, places its purpose outside the realm of technology. (Technology treats various ends only to see whether they are compatible with each other or whether they can be realized.) In this point, it differs from historicism, because historical determinism believes that the ends of human activities depend on the forces of history, so It falls within the purview of historicism. Just as the main task of natural engineering is to design machines and transform and maintain them, the task of progressive social engineering is to design various social constructions and transform and use existing social constructions.The term "social institutions" is used here in a broad sense, including both private and public ones.Therefore, I will use the word to describe a business, whether it is a small store or an insurance company, as well as a school or an "educational system," or a police force, or an a church, or a court.The progressive technologist or engineer realizes that only a few social constructs are consciously designed, and that the vast majority of social constructs simply "grow," the undesigned results of human activity.However, although the incremental engineer may be extremely impressed with this fact, he, as a technologist or engineer, will view social constructs from a "functional" or "instrumental" point of view; means, or that they can be altered to serve certain ends; that they are machines rather than organisms.Of course, this does not mean that he will ignore the fundamental difference between social constructions and practical tools.On the contrary, when the technologist formulates his results in the form of hypotheses, he studies not only their similarities but also their differences.The following example shows that it is indeed not difficult to formulate hypotheses about social construction in technical form. "You can't build a social structure that even a fool can do, that is, its function does not depend on the human structure. At best, you can help people to work in order to achieve the design purpose of the social structure, thereby reducing the disadvantages caused by the human factor. Certainty, success depends heavily on human creativity and knowledge. (Social constructs are like fortresses. "White people must not only be well designed, but also properly staffed.)" Such is the characteristic attitude of the incremental engineer.He may have had some ideals of society as a "whole" (such as the general welfare of society), but he did not believe in a way of redesigning society as a whole.Whatever his purpose, he always takes small adjustments and readjustments that lead to continuous improvement to achieve his purpose.His purpose may be various, for example, the accumulation of wealth or power of certain individuals or groups; or the distribution of wealth and power; or the protection of certain "rights" of individuals or groups, and so on.Thus, public or political social engineering can have a wide variety of leanings, from totalitarian to liberal. (W. Lippmann cites some examples of far-reaching liberal programs of progressive reform under the heading "The Liberal Agenda.") The progressive engineer, such as Socrates, knows how little he knows.He knows we can only learn from our mistakes.He will therefore proceed step by step, carefully comparing the expected results with those achieved, alert to the unavoidable adverse consequences of reform; he will avoid complex and large-scale reforms, because he cannot distinguish the The reasons and consequences of the reform, I don't know what he is doing. This kind of "incremental tinkering" is at odds with the political temperament of many "activists," whose programs have also been described as those of "social engineering," which could be called "holistic" or "utopian engineering." . Holistic or utopian social engineering, as opposed to progressive social engineering, is never of a "private" but always "public" nature.Its object is to transform "the whole of society" according to a definite plan or blueprint; The key position to control the historical forces that affect the future development of society, or hinder the development of society, or foresee its course and make society adapt to it. Some people may ask whether there is any fundamental difference between the incremental method and the holistic method mentioned here, because we have not yet drawn the boundary of the incremental method.So far as it is here understood, constitutional reforms, for example, belong to its sphere; nor do I exclude the possibility that a series of gradual reforms has been caused by some general tendency, for example, towards a more equalization of revenues.In such cases, an incremental approach may lead to a change in what is commonly referred to as the "social class structure".What, one might ask, is the difference between these more ambitious incremental projects and a holistic or utopian approach?The problem becomes more pertinent if we consider that, in estimating the possible results of some reform program, the technologist of progress must try to estimate as accurately as possible the effect of any measure on the "whole" of society. In answering this question, I do not wish to draw a sharp line between the two approaches, but to point out that the holistic technologist and the progressive technologist have quite different views of the task of social reform.Holists object to the incremental approach, arguing that it is too gentle.Their opposition, however, did not correspond to their practice; for in practice, ambitious and ruthless as they were, they were always clumsily applying essentially incremental methods, merely imprudent and lacking in self-criticism.The reason is that the holistic approach is practically impossible: the larger the holistic changes, the more unexpected and highly undesirable their repercussions, forcing the holistic engineer to resort to the expedient measure of incremental improvement .In fact, this expedient measure is more characteristic of central or collectivist planning than of milder and more cautious incremental interventions; it constantly makes the utopian engineer do what he does not want to do; that is to say, it leads Notoriously unplanned plans.In reality, therefore, the difference between utopian engineering and incremental engineering has less to do with scale and scope than with prudence and preparation for the inevitable contingencies.We may also say that, if we compare the two doctrines in their views of the method by which rational reforms should be adopted, they differ not in size and scope, but in other respects, contrary to what we are often supposed to suppose.Of these two doctrines, I think, the one is true and the other is false, and is liable to lead to serious errors which could have been avoided.I think of these two methods, one is possible, the other is simply non-existent, impossible. So one of the differences between a utopian or holistic approach and an incremental approach can be stated as follows: the incremental engineering engineer can ask his own questions without prejudice within the scope of the reform, while the holist cannot. ; because he insisted in advance that a radical transformation was possible and necessary.This fact has profound implications.This makes utopians averse to sociological hypotheses about the limits of control over the construction of society; for example, the one mentioned above in this section that demonstrates the uncertainty caused by the "human factor."The utopian approach rejects these hypotheses a priori, contrary to the principles of the scientific method.On the other hand, the problems associated with the uncertainty of the human factor necessarily compel the Utopian (whether he likes it or not) to use socially constructed means to control the human factor and to expand his programme, To make it not only carry out social transformation according to the plan, but also include the transformation of people. "The political problem is thus to organize human impulses and then to direct them towards the right strategic objectives, so that the general process of development proceeds in the intended direction." This program is open to failure, even before it is put into practice , which seems contrary to well-meaning utopians.Because this program changed his requirements. He originally asked us to build a new society suitable for people's lives, but now instead requires us to "reform" people so that they are suitable for his new society.Obviously, this precludes the possibility of testing the success or failure of this new society.Because for those who do not want to live there, it is enough to admit that they are not suitable to live there, and that their "human impulse" needs to be further "organized".But if there is no possibility of testing, then any claim to scientific method is futile.A holistic approach is incompatible with a true scientific attitude. Utopian engineering is not the main topic of this book, but there are two reasons why it is examined alongside historicism in the next three sections.First, since it is a very popular doctrine in the name of collectivist (or central) planning, "incremental technology" and "incremental engineering" must be sharply separated from it.Second, because utopianism not only resembles historicism in its hostility to incremental methods, but is often associated with historicist ideologies. twenty two.unholy alliance with utopianism The opposition between these two methodological views, which I have termed "incremental technology" and "historicism," was clearly recognized by Mill.He writes, "There are two kinds of sociological research, the first of which asks, . Its question ... is not what is the effect of given causes in certain social formations, but what are the causes of ... social states in general." Considering that Mill's "social formation" is exactly equivalent to our Call it a "historical period," and it is thus clear that the difference between what he calls the "two types of sociological studies" is equivalent to the difference between what we call the progressive technological view and the historicist view.This becomes clearer if we examine Mill's description of the "second sociological investigation" more carefully.He claimed (under Comte's influence) that these two studies were superior to the first, and he said that the second was the application of what he called the "historical method." As indicated earlier (in Sections 1, 17, 18), historicism is not the opposite of 'activism'.Historic sociology can even be interpreted as a technique that can help (as Marx put it) to "shorten and lessen the birth pains of new historical periods".Indeed, in Mill's description of the historical method, we can find that the formulation of this idea is very similar to that of Marx: "The method pointed out here is the method that must be followed in order to explore the laws of social progress. With its help , we can not only look into the future history of mankind, but also decide what artificial means to accelerate this natural progress, so long as it is beneficial... This practical rule based on the highest branch of speculative sociology will become the political art. the most valuable and beneficial part of it." As pointed out here, the difference between my approach and that of the historicists is marked not so much by a technique as by an incremental one.As far as historicism is a technique, its method is not gradual but "holistic". Mill's approach was decidedly holistic when he explained what he meant by "social formations" (or historical periods), writing: "A social formation . . . is all the larger social facts or The shape that phenomena all take." Especially examples of these facts: "The state of industry, the state of wealth and its distribution: the division of society into classes, and the relations of these classes to one another; the common beliefs held by these classes..., Their forms of government, and above all their laws and customs." In summary, Mill characterizes social formations as follows: "Social formations are like ... periods in living beings, which are not the result of one or a few organs or functions. condition, but the condition of the organism as a whole." It is this holism that makes historicism quite distinct from incremental technology, and that makes it possible to ally it with some form of holistic or utopian social engineering. This is of course a somewhat strange alliance; for as we have seen (in Section 15), the method of historicism is not the same as that of a social engineer or technologist if we know how to use social engineering to create social constructs as planned. There is a very clear distinction between methods.From a historicist point of view, the method of historicism is fundamentally opposed to any method of social engineering, just as the method of a meteorologist is fundamentally opposed to that of a rain magician.Consequently, social engineering (even incremental methods) has been slammed as utopian by historicists.Nevertheless, we find that historicism is often associated with typically holistic or utopian ideas of social engineering, such as "blueprints for a new order" or "central planning". The two typical representatives of this combination are Plato and Marx.Plato was a pessimist who believed that all change—or almost all change—is decay; this was his law of historical development.Thus, his utopian blueprint is to prevent all change.This is what people now call "static".Marx, on the other hand, was an optimist who was probably (like Spencer) a follower of historicist morality.Therefore, his utopian blueprint is one of development or "dynamic" blueprint, rather than a bound society.He predicted and actively promoted development with the ultimate goal of an ideal Utopia in which there is no political or economic repression: the state is dead, everyone is free to cooperate according to his ability, and everything that everyone wants is granted. satisfy. The most powerful element in the alliance between historicism and utopianism is undoubtedly that they are both holistic.Historicism is concerned with development, but not with the development of all aspects of social life, but with the development of “society as a whole”; Utopian projects are also holistic.Both ignore the important fact (which we shall present in the next section) that a "whole" in this sense can never be an object of scientific investigation.Neither is content with ''incremental patching'' and ''groping forward''.They want a more drastic approach.Both historicists and utopians seem to have an experience of changing social circumstances (a dire experience sometimes described as "social collapse").I am deeply impressed, and sometimes even deeply disturbed by it.So both of them try to rationalize this change, one is to predict social development, the other is to insist that this change must be strictly and comprehensively controlled, or even stopped completely.Control must be total, because if any part of social life is not fully controlled, there may be hidden dangerous forces for unexpected changes. Another connection between historicists and utopians is that both believe that their purpose is not a matter of choice or moral determination, but that they can discover their purpose by using the scientific method in their field of study. (Here they are as different from progressive technologists or engineers as they are from natural science engineers.) Historicists and Utopians alike believe they can discover the true purpose of "society"; , or assert "the needs of their time," and thus discover the true purpose of society.As such, they tend to adopt a certain historicist moral doctrine (see section 18).It is no accident that most of the authors who advocate utopian "plans" tell us that they are simply inevitable.Because history is moving in a certain direction, we have to plan, whether we want to or not. These authors charge their opponents with intellectual dullness in a tone of historicism, and see their main task as "to overthrow outdated habits of thought and to discover new keys to understanding a changing world." They assert that the changing tendencies of society "cannot be successfully influenced or deflected unless we abandon the incremental approach or the "spirit of groping". However, we can also ask whether that new "planning-level thinking" is as novel, because holism has the characteristics of a fairly old thought that has been around since Plato. Personally, I think we can also find a good example of a holistic way of thinking (whether about "society" or about " Nature") by no means represent a high level or a new stage in the development of thought, but were characteristic of the pre-scientific period. twenty three.Criticism of Holism Having stated my inclinations, outlined the point of view on which my critique is based, and the opposition of the incremental approach to holism and utopianism, I now proceed to my main task, an examination of historicism .I begin with a brief critique of holism, which turns out to be one of the most crucial arguments against historicism to attack. In recent holist writings, the use of the word "whole" is extremely ambiguous.It usually refers to (a) the sum total of all properties or aspects of a thing, especially the sum total of all connections between its constituent parts. (b) Some special property or aspect of the thing that causes the thing to appear as an organic structure rather than a "mere heap." (b) The whole of meaning has been the object of scientific investigation, especially of the so-called Gestalt school of psychology; Regularity can be found in things like organisms, electric fields or machines.Gestalt theory holds that things with this type of structure can be said to be more than collections - "more than the sum of their parts". Any example of Gestaltism will show that (b) the whole of meaning is quite different from the whole of (a).Gestalt theory holds that if we consider the melody of a piece of music to be more than a mere collection or succession of individual sounds, then what we choose to consider is only one of the aspects of this succession of sounds.This aspect can be clearly distinguished from other aspects, such as the first absolute pitch of these speakers or the average absolute pitch of these speakers.There are other gestalt aspects that are more abstract than these aspects of melody, for example, the rhythm of the melody; because when we think of rhythm, we ignore the relative pitch, which is important to the melody.Because of this selectivity, the study of a gestalt, and of any whole in the sense of (b), is distinct from the study of a totality, that is, of a totality in the sense of (a). The fact that (b) the totality of meanings can be scientifically studied does not support the quite different claim that (a) the totality of meanings can also be scientifically studied.The latter claim must be denied.If we want to study a thing.We just have to choose some aspect of it.It is impossible for us to observe or describe the whole world or the whole of nature, in fact not even the smallest whole, since all description must be selective.We might even say that (a) the whole of meaning is never the object of any activity, scientific or otherwise.If we take an organism to another place, we treat it as if it were an object, without paying attention to its many other aspects.If we kill it, we destroy some of its properties, but not all of them.In fact, it is impossible for us to destroy the whole nature of it and the whole relationship between its parts, even if we crush or burn it. The fact that the whole cannot be the object of scientific study, or of other activities such as control or transformation, in the full sense of the term seems to go unnoticed by holists, not even by those who admit that science is often a matter of choice.看到。他们确信科学能够在整体上把握社会（在全部的意义上，因为他们所根据的是格式塔心理学以前的认识。他们相信，格式塔方法和（a）意义的社会整体方法包罗“一个时代的全部社会历史事件的结构”）之间的区别，只是在于格式塔可以直接由直觉感知来把握，而社会整体则“太复杂而不能一下子理解”，“只有经过长期的思索，密切注意其中一切因素并加以比较和综合才能逐渐理解。总而言之，整体主义者没有看到，格式塔的感知同（a）意义的整体毫不相干，他们没有看到一切知识，不管直觉的知识或推沦的知识都必定有抽象的方面，他们也没有看到我们不可能把握“社会实在本身的具体结构”。既然他们看不到这一层，他们就坚持说，专家们对“细节”的研究必须用“合成的”或“综合的”方法来补充，以便重新改造“整个过程”，他们断言，“只要专家们不把他们的问题看作一个整体，社会学就会继续忽视这个本质问题”。然而，整体主义的方法必然只是说说而已。他们对于整个具体社会状况的科学描述从来没有举出一个例子。这样的例子是举不出来的，因为对所举出的任何例子，我们都很容易指出它所忽略的方面，而那些方面在某种场合可能又是重要的。 整体主义者不但企图用不可能的方法来研究我们的社会，而且还企图把我们的社会作为一个整体来控制和改造。他们预言：“国家的权力必须扩大，直到它和社会几乎合而为一为止”。这句话所表达的直觉是够清楚的。这就是极权主义的直觉。这个预言除了转达这种直觉之外还会有什么别的意思呢？“社会”这个词当然包括一切社会关系，包括一切人与人之间的关系；母子关系以及儿童福利官员和母子二者之间的关系等等。有许多理由说明控制所有或“几乎”所有这些关系是不可能的，因为只要对社会关系进行新的控制，我们就创造了一大堆需要加以控制的新的社会关系。简言之，这种不可能性是逻辑的不可能性。（这种作法会导致无穷倒退；其情状与研究整个社会的做法类似，因为对整个社会进行研究就得包括这种研究。）毫无疑问，乌托邦计划正是想去做不可能做到的事情。因为他们向我们所做出的保证之一，就是宣称我们能够“用较为实际的方法来造就人与人交往的形式”。当然没有人怀疑，（b）意义的整体是可以造就、控制或改造的，但对于（a）意义的整体来说则不然；例如，我们可以创作出一个旋律；但这和全面控制的乌托邦梦想毫无共同之处。 关于乌托邦主义就说到这里。就历史决定论而论，这种主张也是没有希望的。历史决定论的整体主义者常常含蓄地断言，历史方法适合于研究在“总体”意义上的整体。但这一断言是由于一种误解而来。它把两种观点混为一谈，其一是正确的观点，即认为历史与理论科学相反，它的兴趣在于具体的个别事件和个别的人，而不在于抽象的普遍规律，其二是错误的观点，即认为历史所关心的“具体的”个人可以视为（a）意义的“具体的”整体。But this is impossible.因为历史和其他学科一样，只能研究对象中被选定的那些有兴趣方面。他们错误地认为可以有整体主义意义上的历史，即表示“整个社会有机体”或“一“个时代的全部社会历史事件”的“社会形态”的历史。这个思想来源于把人类历史看作一个广阔的发展长河这个直观看法。但这样的历史是无法写出来的。每一部写成文字的历史都是这个“全部”发展的某些狭小的方面的历史，总是很不完全的历史，甚至是被选择出来的那个特殊的、不完全的方面的历史。 乌托邦主义和历史决定论的整体主义倾向在如下有典型性的命题上是一致的：“我们决没必要从整体上确定和指挥整个自然界，这与我们今天不得不从整体上探讨我们的社会有所不同。所以我们决没必要深入到历史和自然界的各个领域的结构中去。人类是要……调整全部社会生活的，尽管人类尚未创造出另一个大自然………”这个命题表现出一种错误的想法，认为如果我们象整体主义者那样有意去“完全地研究整个自然界，，，那么采取历史方法将大有帮助。但是，自然科学，例如地理学也采取了这种方法，但远远没有掌握他们课题的“整个系统”。这个命题也表明如下的错误想法：认为我们有可能“确定”或“指挥”或“调整”或“创造”（a）意义的整体。说我们绝没必要确定和指挥整个自然界”这句话，当然是对的。原因很简单，因为我们甚至不能确定和指挥这个“整体”中的一个实际装置。这类事情是不可能做到的。这是乌托邦的梦想或误解。对我们来说，我们现在不得不去做逻辑上不可能的事情，即去确定和指挥整个社会系统，去调整全部社会生活，这不过是用所谓“历史的力量”和“未来的发展”使乌托邦计划成为不可避免之类的话来恐吓我们罢了。 顺便一提，所引用的命题是有意思的，因为它承认一个极其重要的事实：在自然科学中不存在整体主义工程或相应的“科学”的类似情况。所以，进行自然科学和社会科学类比对澄清这个问题肯定是有帮助的。 这就是整体主义的逻辑状况，他们鼓励我们在这种基础上建立了一个新世界。 最后，对意义（b）的整体也批评几句，尽管我承认其科学地位。不必重复我所说过的话，但我必须指出，说整体不仅仅是其部分的总和这句话，既是平凡的真理；但又是意义含混的，这似乎不大为人所认识。甚至一个盘子上的三个苹果也不是“纯粹的总和”，因为它们之间必定有某些关系（最大的那个苹果可能在或可能不在另外两个苹果之间等等）。这些关系并不是从有三个苹果这一事实得出，但可以科学地加以研究。还有，大肆宣传的所谓“原子论”和“格式塔”之间的对立是毫无根据的，至少就原子物理学而论。因为原子物理学并非仅仅是“总计”它的基本粒“子，而是从一个与意义（b）的整体显然有关的观点来研究粒子系统的。 大多数的格式塔理论家显然都愿意说，如下两种情况是存在的：其一是看不出有任何秩序的那些“堆积”，另一是可以发现有某种秩序或对称性或规律性或系统或结构方案的“整体”。因此，诸如“有机体是整体”这样的话就成为很一般的命题，意思是在一个有机体中我们能够发现某种秩序。此外，所谓“堆积”也总有格式塔的方面，例如经常引用的电场（请想一下一堆石子在其压力增加时的规律性）。因此，这个区分不只是平凡，同时又是极其含混；它不能应用于各种不同的事物，而只能用于同一事物的不同方面。 twenty four.整体主义关于社会实验的理论 整体主义思想对历史决定论关于社会实验的理论具有特别有害的影响（上面第2节有所论述）。虽然渐进技术师会同意历史决定论如下观点，大规模的或整个社会的实验即使确有可能，对科学的目的也是极不合适的。渐进技术师将同历史决定论和乌托邦主义一样明确否定如下假定：社会实验要切合实际就必须具有改造整个社会的乌托邦企图的特征。 我们不妨先讨论与乌托邦纲领明显对立的观点，即我们没有从事这种实验所必需的经验知识，并提出我们的批评。自然科学工程师的蓝图是以实验技术为基础的，他的活动所依据的全部原则都受到实验的检验。但是，社会工程师的整体主义蓝图却不是以任何类似的实际经验为基础的。因此，所谓自然科学工程与整体主义社会工程相类似的说法不能成立；把整体主义计划称为“乌托邦”是正确的，因为它的计划根本没有科学根据。 面对这样的批评，乌托邦工程师很可能会承认需要实践经验和实验技术。但是他会宣称，假如我们不去进行社会实验，或者不去进行在他看来与此相仿的事情，不去进行整体主义工程，那么，我们就决不会知道这些事情。他会争辩说，不管我们有多少知识，我们也必须利用我们的知识去做。如果说我们现在有设计飞机的知识，那只是因为没有这种知识的先驱者敢于设计飞机并把它试制出来。因此，乌托邦主义者会争辩说，他所提倡的整体主义方法不过是应用于社会的实验方法。因为，他同历史决定论者一样认为，小规模的实验，例如在一个工厂，一个村庄或一个地区进行的社会主义试验，是无济于事的：这种孤立的“鲁宾逊式的试验”不能使我们了解“大社会”的现代社会生活。这些试验应获得“乌托邦”的绰号——在（马克思主义的）意义上，这个词意味着无视历史趋势。（在这里，其含义就是无视社会生活日益相互依赖的趋势。） 我们看到，乌托邦主义和历史决定论都持这样一个观点，即认为社会实验（假如有这样的事情的话），只有以整体主义的规律来进行才有价值。这种普遍抱有的偏见包含着一种信念，即认为我们没有多少可能在社会领域中实行“有计划的实验”，而且考虑到迄今在社会领域中所进行的“机遇实验”的结果，我们不得不把注意力转向历史。 我对这个观点有两点反对意见：（a）它忽略了对于一切社会知识，包括前科学的和科学的知识都至为重要的渐进实验。（b）整体主义实验不可能对我们的实验知识有较大的帮助；它们只能在与“其结果未知的行为”同义的情况下称为“实验”，但这个词通常用来表示把结果和所预期的结果相比较的一种获得知识的方法，在这个意义上，整体主义实验就不能被称为“实验”。 关于（a），可以指出，整体主义关于社会实验的观点尚未解释我们有着许多关于社会生活随实验知识这一个事实。有经验的和无经验的实业家、组织者、政治家或将军等，是不相同的；他们的不同表现在社会经验上：而经验的获得不仅通过观察，或对他们所观察的事物深入思考，而且要靠为达到某种实际目的而付出努力。必须承认用这种方法所取得的知识通常属于前科学的类型，因此它与其说是从精心设计的科学试验中所获得的知识，不如说是从偶然观察中所获得的知识；但是，没有理由否认这种知识是基于实验而不仅仅基于观察。一个杂货商人开一家新商店就是进行着一种社会实验；甚至一个在剧场门前排队的人也会获得实验性的技术知识，他可以在下次排队时利用这一知识来定座位，而这只是一种社会实验。我们不应忘记，只有实际的尝试才使市场上的卖主和买主认识到价格会随供应的增加而降低，并随需求的增加而升高。 规模稍大的渐进实验的例子是垄断者决定改变他的产品价格，私人的或公家的保险公司开办新的健康保险或职业保险业务，或设立新的销售税，或实行反对经济周期的政策。所有这些实验都是着眼于实践的目的，而不是着眼于科学的目的来进行的；某些大公司所进行的实验则是特意为增长他们的市场知识（当然是为了以后增加利润）而不是为了马上增加利润。这和自然工程以及前科学方法的情形十分相似。我们的技术知识，例如造船或航海方面的知识，最早是通过前科学的方法获得的。看来，我们没有理由说这些方法不必加以改进井最终让位给较注重科学的技术；这就是说，让位给以批判性思维和以实验为基础并朝着同一方向的较为系统的方法。 按这种渐进观点，在前科学方法与科学实验方法之间并无明确的界限，尽管日益自觉运用科学的（即批判的）方法是非常重要的。从根本上说，两种方法都可以说成是运用。我们做试验，即我们不仅仅记录我们的观察结果，而是主动地试图去解决某个多少实际的和明确的问题。当且仅当我们准备从我们的错误中学习，认识到我们的错误并批判地加以利用，而不是固执地坚持错误，我们就会取得进步。虽然这个分析听起来很平常，但我相信，它说明了一切经验科学的方法，这种方法越来越具有科学的特征，我们越自由而自觉地准备去冒试验的风险，我们就越能够以批判的眼光去找出我们总是犯的错误，这个公式不仅包含实验的方法，也包含理论与试验之间的联系。一切理论都是尝试，都是试验性的假说，它们是否成立都要经过检验，而一切实验的确认则不过是以批判精神进行试验的结果，为努力发现我们理论的错误而进行试验的结果。 对于渐进技术师或工程师来说，这些观点意味着：如果他希望在社会和政法的研究中采用科学方法，那么，最必要的是采取一种批判的态度，并且认识到做尝试和犯错误都是免不了的。他不仅必须学会预料到会出错，而且必须有意识地去寻找错误之所在。我们所有的人都有自认一贯正确的违反科学的弱点，而这个弱点似乎在职业的或业余的政治家当中尤其普遍。然而，在政治学中应用某种科学方法的唯一途径就是首先认定，有政治运动就会有缺点，就会出现不希望有的结果。警惕这些错误，找出它们，把它们公开出来，进行分析，并从这些错误中学习，这是一个有科学眼光的政治家和政治科学家所必须做的。政治学中的科学方法意味着那种确信我们没有犯任何错误，无视错误，掩饰错误，或把错误归罪于人的伟大艺术让位给另一种更伟大的艺术——为错误承担责任，力图从错误中学习并应用这一知识避免将来犯错误。 现在我们转到（b），即对那种认为我们能够从整体主义的实验中学习，或者更确切他说，认为我们能够从以整体主义梦想为目的所实行的大规模措施中学习的论点给予批评（因为，正如我在前一节里所表明的，整体主义的实验在其极端的意义上是要重新建造“整个社会”，这在逻辑上是不可能的）。我们的主要论点很简单，这就是：对我们自己的错误采取批判的态度确实非常困难，而我们对自己所采取的涉及许多人的生命的行动始终采取批判的态度就必定几乎不可能。换句话说，从重大的错误中学习是很难的。 其中的原因有二，它们既是技术上的，又是道德上的。既然在一个时期里做那么多的事情，那就不可能指出哪一种措施应对某一个结果负责，或者更确切他说，假使我们把某一特定结果归因于某一特定的措施，我们也只能以从前所取得的某种理论知识为根据，而不能从该整体主义实验中知道。这种实验不能帮助我们把特定的结果归因于特定的措施；我们所能做到的只是把“总的结果”归因于它；而且，不论这可能意味着什么，要评论它确实很困难。即使尽最大的努力对这些结果提出结构完善的、独立的批判性陈述，也不能证明能够成立。况且，进行这种努力的机会是微乎其微的，相反，对整体主义计划及其结果的自由讨论很可能是不被容忍的。原因是实施极大规模计划的尝试是一件使许多人至少在相当长的时期内感到十分不便的事情。因此，总会出现反对和抱怨的倾向。对于许多抱怨，乌托邦工程师如果还想继续干下去，那就只好充耳不闻。事实上，压制不合理的反对意见是他的工作的一部份，但他必定也不断地压制合理的批评。光是对不满言论进行限制这个事实，就会使最热烈的满意言论变得毫无意义。因此，这就很难确认各个公民对这个计划如何反应等事实，而没有这些事实，就不可能有科学的批评。 然而，把整体主义计划与科学方法结合起来的困难，比我们迄今所表明的还要严重得多。整体主义计划者忽略了这样一个事实：权力集中是容易的，但是把分散在许多人头脑中的知识集中起来是不可能的，可是这种集中对于明智地运用中央集权是必要的。这一事实具有深远的意义，既然不能确定在这么多的人的头脑中的想法，他就只好消除个人之间的差别而使自己的问题简单化：他必须用教育和宣传来控制和统一人们的兴趣与信念。然而，这种试图控制人们的精神的做法，势必会毁掉发现人们真正思想的最后的可能性，因为它显然与思想自由，特别是批判思想的自由不相容。其结果，它必然毁掉知识；权力越大，知识的损失也越大。（由此可以发现，政治权力与社会知识是玻尔所说的“互补的”。很可能这是对这个难以捉摸而又很时髦的词的唯一清楚的说明。 所有这些话语只限于科学方法的问题。这些话无形中承认如下的重大假定：我们不必追问乌托邦计划工程师心底里的仁慈心，总之他们至少有取得独裁权力的权威。托尼（Towney）对马丁.路德和他的时代的讨论得出了如下的结论：“尽管不相信有独角兽和火怪，但马基雅维里（Machiavelli）和亨利八世的时代仍为崇拜那个稀有怪物——害怕上帝的君主的盲信找到了支持。”这里是用“害怕上帝的君主”替换了“独角兽和火怪”，把这两个名称替换为两个较明显的现代对应物，把“害怕上帝的君主”替换为，'仁慈的计划权威”。于是，我们就有了关于我们时代的盲信的描述。在这里，我将不反驳这种盲信。然而，我们可以说，即使有权势的计划者有着无限的、始终如一的仁慈，我们的分析也会表明，他们绝不可能得知他们的措施的结果是否与他们的良好愿望相符合。 我不相信能对渐进方法提出适当批评。这个方法尤其可以用来找寻和克服最严重和最迫切的社会弊病，但不是去寻求某种终极的善并为之奋斗（这是整体主义所要做的）。采取有步骤的措施来反对某些错事，反对不公正或剥削等具体情况，反对可以避免的苦难（例如贫困和失业），与试图实现一个遥远的理想社会蓝图截然不同。成功或失败的鉴定是比较容易的，没有内在的理由说这种方法会导致权力集中和压制批评。反对具体的错误和具体的危险比起力图实现计划者认为十分理想的乌托邦来，更能得到绝大多数人的支持。这也许能够说明如下事实：在反侵略的民主国家里，必要的长远措施（这些措施可能带有整体主义计划的性质），会获得充分的支持，而无需压制公众的批评，然而，在准备发动进攻或侵略战争的国家里，公众的批评总是受到压制，甚至为了动员公众的支持，硬把侵略说成自卫。 现在我们回到乌托邦主义者的论点，他说自己的方法是应用于社会学领域的真正的实验方法。我认为这个论点已被我们的批评所驳倒。这一点可以用自然工程与整体主义工程之间的类比作进一步的说明。可以认为，机器可惜助蓝图成功地设计出来，不仅如此，甚至生产机器的整个工厂也可以借助蓝图成功地设计出来。这一切之所以可能，只是因为事先已进行过许许多多的渐进实验。每一台机器都是许许多多的小改进的结果。每个模型必须经过反复实验，经过无数次的小调整而“发展”起来的。生产厂的计划也是如此。显然，整体主义计划之所以能够成功，仅仅因为我们已经犯过各种各样的小错误，如果不是这样的话，我们有理由设想它会导致大错误。 于是，自然工程和社会工程之间的类比，深究起来，不利于整体主义的社会工程师，而有利于渐进的社会工程师。包含这个类比的“社会工程”这个词已被乌托邦主义者毫无理由地盗用了。 至此，我结束我对乌托邦主义的批评，并将集中抨击它的同盟——历史决定论。我相信我已经对历史决定论者关于社会实验的论点给出了充分的回答，而只剩下如下的论点：说什么社会实验是无用的，因为在完全相同的条件下进行重复的实验是不可能的，我们现在就对这个论点加以考察。 25．实验条件的可变性 历史决定论者争辩说，实验的方法不能应用于社会科学，因为在社会科学领域里，我们不能随意再现完全相同的实验条件。这使我们稍为接触到历史决定论观点的核心。我承认，这个主张是有点根据的，因为在社会科学方法与自然科学方法之间无疑是有些不同的。然而，我还是认为，历史决定论的主张是基于对物理实验方法的严重误解。 让我们先来考察这些方法。任何一位实验物理学家都知道在看来完全相同的条件下会发生极不相同的情况。两条电线，乍看起来完全相同，但是如果在一台电器设备中加以互换，就会有极不相同的结果。在更精细的检查下（例如通过显微镜），我们就会发现它们并不象原来所设想的那样相似。但是要察觉到导致不同结果的两个实验的条件之间的差别，确实往往是很困难的。需要进行长期的实验研究和理论研究才能找到究竟哪种相似性是相关的，以及其间相似到何种程度才是足够的。我们必须进行这种研究，然后才能确知实验所要求的相似条件，甚至才能知道在这种情形下“相似条件”是什么意思。尽管如此，实验的方法在任何时候仍然被采用。 因此我们可以说，什么叫做“相似条件”这个问题，取决于实验的种类，并且只能运用实验来解答。被观察到的异同无论怎样显著，先验地判定其相异和相似，判定与再现实验是否相关，是不可能的。因此，我们必须让实验方法自己解决自己的问题。严格相似的说法是针对排除干扰作用的人工隔离实验这个议论纷纷的问题而提出来的。很明显，我们不能使一个装置与一切影响相隔离，例如，我们不可能先验地知道各个行星或月亮的位置对一项物理实验的影响究竟是值得考虑还是可以忽略。我们只能从实验的结果或者从经过实验检验的理论中知道究竟哪种人工隔离是必需的。 历史决定论的论点认为，社会实验受到社会条件的可变性的致命限制，特别是受到历史发展所引起的变化的致命限制：鉴于这些考虑，历史决定论的这一论点缺少说服力。事实上，历史决定论者十分注重的显著的区别，即各个历史时期的一般条件之间的区别，不一定给社会科学造成特殊的困难。可以承认，假如我们突然被转移到另外一个历史时期中去，我们可能发现，我们进行的渐进实验所产生的许许多多社会期望全部落空了。换句话说，实验可能导致预想不到的结果。然而，正是实验使我们发现社会条件的变化。实验会告诉我们，一定的社会条件是随着历史时期的改变而改变的；就象实验告诉物理学家开水的温度是随着地理位置的不同而不同一样。换句话说，关于各个历史时期之间的区别的学说，不会使社会试验成为不可能，而仅仅是一种设想的表述，即认为我们如果转入另一个时期，我们还会继续进行渐进的实验，只是会达到意料不到的或失望的结果。事实上，如果我们知道对不同的历史时期要采取不同的看法的话，那么，这也是来自我们在想象中所进行的实验。历史学家在解释某些记载时碰到困难，或者他们发现一些事实表明他们的前辈曾错误地解释某个历史证据。这些历史解释的困难是历史决定论者心目中的历史变化的唯一证据。然而，这些困难只不过是在我们的思想实验中预期的结果和实际的结果之间的矛盾罢了。正是这些意料不到的、令人失望的结果，借助反复试验法，提高了我们解释新遇到的社会条件的能力。我们在历史解释中通过思考实验所取得的东西，已经由人类学者在实际领域的努力中取得了。现代的研究工作者们已成功地使他们的期望切合于石器时代那么遥远的条件，他们的成功归功于渐进的实验。 有些历史决定论者不相信这种成功调整的可能性；他们甚至为了捍卫社会实验无用论而反驳说，如果回到遥远的历史时期去，那么落空的社会实验就会多得多。他们认为，我们不可能使我们的思想习惯，特别是我们分析社会事件的习惯，来适应那些难以理解的条件。我认为这些担心在某种程度上是历史决定论者神经过敏——对社会变化的重要性着了迷；但我又必须承认克服这些先验的担心是困难的。毕竟，使自己适应新环境的能力，是因人而异的；我们似乎没有理由期望历史决定论者（他持有这种失败主义的观点）能够成功地使自己的思想切合于社会环境的变化。再者，问题还有赖于新环境的特点。不能排除社会研究者在他还没有通过反复试验成功地使自己适应于人吃人的习惯之前就被人吃悼的可能性，同样也不能排除在“计划社会”中他的研究以自己被关入集中营而告终的可能性。然而，类似的说法在自然科学领域中也是成立的。在许多情况中，自然条件是主要的，这使自然科学家没有多少机会通过反复试验来使自己适应这些条件。 总之，似乎没有什么根据支持历史决定论者的似是而非的论断，说历史条件的可变性使实验方法不能应用于社会问题，说社会的研究在这一点上根本不同于自然科学的研究。至于我们承认社会学家实际上难以随意选择和改变他的实验条件，那是不同的问题。自然科学家的处境要好一些，但他有时也碰到类似的困难。在变动的引力场或者在极高或极低的温度条件下进行实验的可能性是十分有限的。但是我们千万不能忘记，今天向自然科学家提供的许多可能性，在不久之前还是行不通的，这不是因为自然界的困难，而是由于社会的困难，即因为我们不曾准备花钱去研究。然而事实上，现在许多自然科学的研究能够在不需要多大改善实验条件下进行，而社会科学家的处境则全不同，社会科学家极想做的许多实验在今后很长的时期内仍然是一些梦想，尽管它们不是乌托邦的实验而是渐进的实验。事实上，他不得不常常依靠在头脑中进行的实验，依靠在从科学的观点看来尚待大大改进的条件下对政治措施进行分析。 26．概括局限于时期吗？ 我先讨论社会实验的问题，然后才详细讨论社会学的规律、理论、假说或“概括”。这并不意味着我认为观察与实验在某种意义上在逻辑上先于理论。相反，我认为理论先于观察，也先于实验，因为观察与实验只有与理论问题相联系时才有意义。而且，我们必须首先有了问题，然后才能希望观察和实验能帮助我们提出答案。或者，就反复尝试法来说，实验必定先于错误的出现；而且，我们知道（在24节），理论或假说是试验性的，它是实验中的一部分，而观察与实验由于表明理论在什么地方出错而帮助我们把它们淘汰。因此，我不相信“概括的方法”就是说我不相信科学开始于观察并通过概括或归纳的过程从观察得出科学的理论。我却认为，观察与实验在帮助我们检验理论和淘汰经不住检验的理论方面所起的作用是比较有限的，尽管我们必须承认这个淘汰过程不仅仅核查理论思考，同时也促使理论思考再去尝试，——往往是再犯错误，再次被新的观察和实验所驳倒。 在本节中，我将批评历史决定论如下观点：它认为社会学中一切概括的有效性，或者至少是最重要的概括的有效性，只局限于进行相关观察的那个具体历史时期之内。我没有先讨论所谓“概括方法”是否成立这个问题之前，就批评这个论点，尽管我相信这个问题并不成立：因为我认为，我无需表明这个方法无效就能把历史决定论这个论点驳倒。所以，关于我对这个方法以及关于理论与实验之间的一般联系的看法可以暂时搁置，我将在第28节中再提出来讨论。 在我对历史决定论者的论点进行批判时，我首先承认处在特定历史时期的多数人都会错误地认为，他们在周围所观察到的规律性都是社会生活的普遍规律，因此对任何社会都适用。事实上，当我们在外国时，我们发现我们对待食物的习惯，我们的礼仪等等，并不是象我们天真地认为那样是可以被接受的，这时我们才注意到我们抱有这些信念。显然可以推知，我们有许多其他的概括，不管我们是否有意识地持有它们，都可能属于这一类，它们可能尚未受到挑战，因为我们不可能到另