The Case for a Formal Conception of Economics

Claudio Gutierrez

Let us briefly review what we have accomplished up to this point in this study. We began by stating in the opening chapter our acceptance of a conception of science that makes it ultimately dependent upon intellectual passions and values to which the scientists are committed. We also pointed to the important truth that it is impossible to make a sharp distinction between pure theory and empirical or historical happening. An interaction always exists between these two poles of knowledge. We further emphasized the fact that value is always encountered in interrelation, forming systems, that some system of value always exists, and that different systems dwelt in by different persons are most likely to conflict with one another. The upshot of this recognition of conflict was not, however, a plea for relativism, but rather for tolerance and hope, hope that conflicting views may eventually be reconciled by means of mutual respect and intellectual, heuristically creative, discussion. This openness means also, given the necessary concomitance of knowledge and habit, the disposition to work with, and even be worked upon intellectually by, our fellows, which can imply painful processes of affiliation and even struggle. But all this is conducted under a firmament of valid, even if revisable, shared–at least to a point– beliefs and loyalties to values. The objective social cosmos is the only possible ground for the validity of sense observation and of scientific reasoning themselves.

We went on to examine the claim that there is an essential distinction between commonsense and science, and that in this distinction the difference between the natural and the social sciences could be founded. We were able to see that some writers consider the social sciences pejoratively as only commonsense knowledge, whereas some others consider it to be an asset rather than a defect to be the direct heirs of prescientific knowledge. We concluded from that examination that the important distinction is not that between commonsense and scientific knowledge, but rather the distinction between formal and non-formal knowledge, between formalism and the unformalizable basis needed for the meaningfulness of it all. We came to the conclusion that it is impossible to objectively found objectivity, to formally support formalism, since that pretension could only lead us into infinite regression. The only reassurance we can expect for our need of ultimate epistemological security is the confidence in the responsibility of the scientist who is trying to fulfill standards of universal intent which he himself sets for his own performance. (POLANYI 64, p. 214). That responsibility, coupled with an efficient development and use of the devices of formalism, is what we are entitled to call the professional sense of the practitioners of science.

In this context, we developed what could be considered a central idea of this essay, the distinction between two types of teleology or subjectivity: teleology of the explanation and teleology in the subject-matter. NOTE 1 In connection with this distinction we construed what we called the temptation of the social scientist, i.e., to identify with each other two levels of analysis that must be kept conceptually separate. The reflexivity implied in the fact that his subject-matter is as subjective as the social scientist himself has led the methodologist into believing that an altogether special scientific logic must be available to the social scientist which is not the common patrimony of the practitioners of general science. This temptation he should do well in resisting, being conducive only to intellectual confusion. It is enough for all scientific purposes to treat subjectivity in-the-subject-matter by means of theoretical terminology, as it is done with "unobservables" in all domains of science.

Closely connected with the issue of teleology and subjectivity is the problem of the nature and uses of abstraction. I tried to react, in the chapter dealing with this subject, against a conception of abstraction which rests upon an unidimensional interpretation of reality. In that part of this study I took pains to show that the abstractionist's "analytical realism" is an insufficient account of the multidimensionality of the real world. I did it especially by exposing an underlying Gödelian overtone in Parsons' epistemology, and branding it as a more nearly correct insight than the "conjunctive" interpretation of the methodological proposal of that writer. The conception that all theoretical thinking is totalizing and systematic, and at the same time somehow limited or not complete enough to render a full account of reality, is the reencounter in the methodological realm of the fundamental truth of the interrelatedness of value and the conflicting character of the relationships between different systems of value. Again, there is here a case of universality vs. exclusiveness. But now we find it at the core of reality, or better perhaps, in the very means of our interpretation of reality. Conflict is not only ever present; it is ineradicable in the single mind, in its very capabilities for the understanding and generalizing about reality. Ambiguity, we are led to say, is inevitably inherent in all thinking bearing on reality, and we cannot hope to directly dissolve that ambiguity in order to "see through" it to the "real world". Our way to truth must be much more complicated than the simple "reading-off" of the data of sense experience or of some inscrutable inner intuitions. The way to truth, I propose, must be related to the fortunate fact that the ineradicable ambiguity we have to recognize is not random but systematic. It makes us be concerned with the thorough examination of competing comprehensive views or paradigms. The adjudication among them must be connected with their coherence or harmoniousness, together with the fullness of account that they are capable of bringing about diverse elements of reality. Circularity of "long range" is implied, but this kind of circularity is not a defect in ultimate knowledge. Nevertheless, it is still true that some internal criteria must be discovered to tell apart "long" from "short" circularity, once the paradigms have passed the all-important test of consistency.

The idea of paradigm was then further discussed. Paradigms are alternative, necessarily circular, comprehensive accounts of empirical appearances, competing interpretations of reality, so to speak. They are eminently subjective, in the sense that they are asserted or dwelt in by persons committed to their eventual truth. But they are universal in intent in the sense that it is to their truth to what the persons entertaining them are committed. Truth and falsity of a paradigm cannot be separated, in ultimate analysis, from the belief or unbelief of the persons discussing the content of the paradigm. A criterion of rational stability of belief comes to be also the criterion of truth, and this is ultimate coherence. The property of making ultimate coherence or sense, however, is not unlimited. It does, in fact, become exhausted. The property of exhaustion or depletion of a paradigm, or of a set of them, might be regarded as a fundamental psychological property of the person asserting the paradigm.

At the empirical end, the capacity of the person for making changes in the paradigm in order to save it from adverse evidence may become exhausted. The paradigm must then be dropped as false. At the theoretical level, the ability of the person to imagine alternative frameworks of explanation can also become depleted, and the paradigm must be asserted as necessarily true. But these two points of exhaustion are ideal poles, never fully attainable, like complementary foci only asymptotically approached. NOTE 2 Within these two poles the concrete instances of knowledge do occur, as it were, in a state of flux from the (ideal) necessary theory to the (ideal) empirical fact, as an indefinite gradation of pieces of information that are both theoretical and empirical, according to the role they play in relation to other pieces of information. In general, one can say that a piece of knowledge is a model, a close enough representation of empirical reality, if one works on it from a higher level of knowledge. On the other hand, a piece of information is an assumption, a theory or paradigmatic configuration, if one dwells in it in order to work from it on something else. Within a belt of knowledge sufficiently distant both from "necessary assumptions" and "overwhelming evidence" there is much leeway to alternative organization of models and assumptions. Many methods might appear to be intellectually fruitful, several approaches equally sound and many conflicting hypotheses equally true, provided the necessary correlations within the paradigm in regard to coherence are always taken care of.

Among the most important of these correlations are those needed for the logical closure of the particular system of hypotheses. This requirement stems from the fact that logical or formal considerations always calls for a cutting-off of non formal context. If the living body of knowledge–or belief–can be visualized as a picture in depth, the formal version of a segment of it has to be visualized as a plane within that space. The contextual depth is represented within the flat logical picture, like a sort of projection, by residual categories, formal tokens of the non formal indispensable assumptions. Thus, empirical openness, necessary for the true interpretation of the system, and logical closure, necessary for the effective formal operation of it, are both concomitantly assured.

Now, under this tolerant methodology one may wonder what one could possibly say against anybody's position, not clearly inconsistent, provided it does not clash with "necessary assumptions" or "overwhelming evidence," which anyhow are very difficult to characterize. To this I must answer in two ways. The first is touché! It is certainly very easy for this methodology to degenerate into "jelly" epistemology, making room for anything. Again, we must rely on the sense of responsibility of the scientist, in this case the methodologist or the philosopher, to avoid this pitfall. The second answer is this. A theoretical view that dissolves a problem rather than solving it, provided of course that the professional considers it a real problem, is not broad enough, not explanatory enough, and should not pass the test for completeness even for a tolerant methodology. NOTE 3 Furthermore, one can say that this epistemology is the reply, and I think the only valid reply, to a literalistic conception of economic theory, and for that matter to any literalistic conception of scientific theory generally. By "literalistic" I understand the property of a methodological system that is most conspicuous in aprioristic epistemological explanations, e.g., Von Mises' praxeology, but which is also in evidence in many middle-of-the-road positivist methodologists, like Robbins, Hayek, or Friedman. I am referring to epistemological theories which conceive of language as something given and inert, incapable of revision or of any dialectic movement of its own, either a filing system unaffected by the content of the file or a knowledge-of-our-own-mind not enriched by intercourse with experience. Furthermore, I consider literalistic a system of thought that conceives of reality as essentially exhaustible in terms of human knowledge, in an unidimensional or extensional sense; i.e., a system which makes no essential room for conflict and ambiguity in the interpretation of reality, postulating as a possible conception the (real or asymptotical) acquisition of a body of total beliefs simultaneously consistent and complete. I have already shown that the gradualist methodology is an effective weapon in the criticism of these positions. In the chapter on Verstehen I attacked literalism in the methodology of the social sciences generally, and in the chapter on praxeology I attacked literalism in the methodology of economics particularly. Those were logical criticisms. In a way, they were easy, once the general gradualistic approach was set up. A constructive presentation of alternative explanations may prove more difficult. But I think it is also necessary and, of course, possible. It would be the demonstration of the actual feasibility of diverse models of quite general type where the literalistic position will maintain that there is only one valid representation of economic reality. To this task we must now turn. It will imply an endeavor of "contextual formalization" of the kind that was described in the preceding paragraph.