catallaxy files

catallaxy in technical exile

Jack Birner on explanation in the human sciences

leave a comment »

Over at Oysterium I am running a series on sociology and economics, starting with a critial appraisal of the remarkable contribution of the late Talcott Parsons. I have also introduced Jack Birner as one of the most helpful people in this area and I want to put up a long piece describing one of his papers but they don’t appear to have folds on Oysterium so I have decided to put it up here instead. Lucky Catallaxians!

Obviously this will have interest to some of the folk here but I recognise that it is not the kind of thing that turns everyone on.

Jack Birner has written a paper titled “A roundabout solution to a fundamental problem in Menger’s methodology and beyond”. This is in a collection on Carl Menger’s legacy printed as a supplement to vol 22 of the “History of Political Economy” in 1990 edited by Bruce Caldwell. Birner explains how some of Carl Menger’s ideas might be brought up to date to make a significant contribution to the epistemology and methodology of the social sciences.

Menger was one of the big three names in the marginalist revolution (with Jevons and Walras) and the founder of the so-called Austrian tradition in economics and social science. He is probably best known for his fiery confrontation wtih the historical school led by Schmoller in an exchange known as the “methodenstsreit”.

Menger published two books. The first was Principles of Economics (1871) which was his contribution to the “marginal revolution” in economics. The key insight of the marginal revolution was the subjective theory of value, whereby the value of a good is decided by the amount that someone is prepared to pay for it, not the amount of labour that went into making it.

The significance of the term “marginal” is that the subjective value of a good (among other options) declines as we have more of it. According to Jack Birner’s account, Menger’s first book was written in diplomatic terms and was dedicated to Wilhelm Roscher, the protagonist of the older Historical School. Menger offered a complementary approach to the historical school, with the implication that different problems required different methods. In particular Menger defended the idea that there are economic laws that apply to all times and places, and it is the search for these laws that constitutes the taska of theoretical economics, as opposed to historical studies and technnological applications. The Historical School refused to accept the “Manchester” doctrine of classical economics, in favour of the idea that every period had its unique character so that there are no general principles of economics or social science to be found.

Menger’s second book came over a decade after the first. Investigations into the method of the Social Sciences (1883) was more strongly worded in order to deliberately start a debate about methodology. He felt that his first book had been ignored because the Historical School did not accept the idea of general principles. He was not opposed to historical studies and he wanted to sketch a bigger picture to show how the historical sciences and the generalising sciences could coexist and cooperate instead of talking past each other.

Birner suggests that the chief message of the book is a strong methodological pluralism – “To each goal its method”. Menger identified various subdisciplines which he considered could be found in both the natural and the human sciences.

1. Historical studies of individual or concrete phenomena and sequences of events.
2. Morphological sciences, which pave the way for the theoretical sciences.
3. Theoretical sciences which yield general laws of typical phenomena.
4. Practical sciences or technologies.

Menger wanted to explicate the methodological rules that would permit economics to qualify as a theoretical science (type 3) and Birner suggests that he should have been more clear about this by stating that “the exact theoretical sciences are about certain sides of all phenonmena in abstraction from disturbing factors”. Birner wrote “Most methodological precepts have their roots in ontological considerations, and Menger’s are no exception. He believes in the existence, in reality, of several fundamental motives for human behavior, which he calls alternatively ‘influences’, ‘goals’, ‘forces’, ‘drives’, and ‘fundamental tendencies of human nature’.”

These are the economic drive, moral sentiments, altruism and justice. The influences of these drives are (respectively) the subject matter of economics, social philosophy, ethics and jurisprudence.

Each of these influences are subject to disturbing influences, among them error, ignorance, force and neglect.

So any particular action can have mixed motivation and it can also be diverted from (or be inappropriate to) its prima facie object by one or more of the disturbing factors.

This means that the laws of supply and demand are not falsified by charity or by paying a high price for bread that is cheaper in the next shop, any more than the law of gravitational attraction is falsified by an apple that is firmly attached to the tree.

Every science has to develop methods to take account of alternative influences and disturbing factors, other than the ones that you are working on.

Birner then went on to examine the logical and epistemological problems that Menger encountered. He was a scientific realist and he maintained that both empirical and exact theories are descriptions of reality. But he ran into problems with justificationism.

For Menger (and his contemporaries, with the possible exception of Whewell) the logical or epistemological problem of the relation between exact and empirical theory is a problem about the justification of knowledge: [how can knowledge] be given a foundation that is true beyond doubt?] At this point Menger encountered the problem of induction, which was supposed to be the method for obtaining foundational knowledge about the world.

Menger’s joint justificationist-inductivist theory of knowlege entails that abstraction is conceived of as a process rather than as the description of a set of hypotheses with particular properties, regardless of how they were arrived at. But Menger is not a naive inductivist. He is well aware of the logical problem that arises if one maintains that general, universally valid laws can be derived from a finite number of observation statements.

Menger then turned to the construction of “pure types” as a way out of the dilemma but he never broke out of the inductivist framework. This may be why he published so little in the last 30 years of his life. Weber took up Menger’s ideas and developed the methodology of idea types, but Birner leapt over that chapter to take up Popper’s more recently developed ideas on situational analysis which is the modern version of “ideal typology”.

“I shall discuss Popper’s analysis of theoretical explanation in the social sciences not only because of its connection with Menger’s problem, but also because the problem is still, more than a century after it was first noticed, real, relevant, and unsolved”.

Birner worked through Popper’s paper on the rationality principle and situational analysis to address the issues of laws in social science, their content, their status and their truth.

Under the heading “Are there laws of social science?” he wrote:

Popper is a methodological individualist; thus, if there are laws of social science, they must be, or be reducible to, laws describing the behavior of individuals. Popper states that there is one such law, which he dubs the rationality principle. It says that people act in accordance with their situation.

As Birner commented, the situation bears a great deal of the weight of explanation, and it is no simple thing because, as Popper used the term in this context, it contains all the relevant aims of the actors and all the available relevant knowledge including that of the means for achieving their aims. Birner went on “The rationality principle, however, is not to be conceived of as subjective or psychological; given the situation, there is only one rational course of action. Latsis called this type of explanation ‘single -exit explanations'”.

[Comment (a) It seems that Popper was very concerned to avoid any whiff of psychologism or subjectivism. He wanted a rational and objective explanation of human action (even allowing for error) and in his desire to avoid psychlogism he may have overlooked a way of giving a non-psychological account of subjective appraisals and decisions. This comment anticipates the ideas of Koertge advanced by Birner later in his article

Comment (b). It may be that the desire for a single-exit explanation manifests a residue of justificationism, that is, the notion that we should be able to come up with a “right answer” (a justified true belief ) that represents the proper solution to the problem situation. The alternative, non-justificationist approach accepts that the actor’s decision is irretrievably conjectural.]

For Popper the assumption of full rationality and perfect knowledge constitutes the “null hypothesis” and one would then investigate the factors in the situation that account for deviations from that ideal behaviour. Menger provided a longer list of idealised assumptions [in the form of the various motivations and the distorting factors noted in the previous post. To recapitulate, the motives are economic drive, moral sentiments, altruism and justice. The disturbing influences are error, ignorance, force and neglect (not fucussing fully on the economic considerations)].

Birner then proceeded to consider “What is the methodological status of the law of social science?”

For Popper the rationality principle has the status of a fundamental law in physics, like Newton’s laws of motion, it serves to set the model in motion, to “animate” the social situation. Popper wrote “We need, In order to animate it [the model of the social situation] no more than the assumption that the various persons or agents involved act adequately or appropriately; that is to say, in accordance with the situation.”

Similarly, Menger considered that the “definite direction of will on the part of the acting invididual” gave explanations in the social sciences a formal analogy with explanations in the natural sciences.

Moving on to the question “Has the law empirical content, and is it true?” Birner noted that for Popper, models based on the rationality principle may be tested (though the principle itself is almost empty) but for Menger the notion of testing the exact theory of economics against full empirical reality is absurd (due to the large number of complicating factors). As for the truth status of the law, both Popper and Menger deny that it is a priori valid or that it is derived by conceptual analysis.

[Popper’s treatment of this topic is probably the most confused part of his work, though it may not be too difficult to fix it up using some of his ideas, plus some Austrian insights and some thoughts from Noretta Koertge and Birner himself].

Birner wrote:

There are several inconsistencies in Popper’s account of the rationality principle. The RP is a methodological principle, yet it is empirical; it has little empirical content, and it is false, thought a good approximation to the truth; but despite its falsity it must not be rejected or replaced by a principle that is closer to the truth.

Both authors run into problems, and neither solves them: Popper because he lacks an articulated metatheory of the character of idealizing theories, and Menger – although he stresses the idealizing character of theories – because his justificationist epistemology prevents him from explaining how theories on different levels of abstraction are related.

Birner then went on to describe Koertge’s contribution. This consisted of increasing the empirical content of the rationality principle, by injecting more detail into the description of the situation. My gloss on this is that “rationality” in the abstract is replaced by the Austrian notion of purpose in a framework of rules, laws, and traditions with further constraints provided by individual dispositions and perceptions. Depending on the nature of the situation and the time available, the actor can take in more or less of the situation, calculating more posible options and consequences before making a move. The recreation of the situation is inevitably conjectural but that is not a problem for non-justificationists.

It is a little surprising that neither Koertge nor Birner conjured up the shade of Max Weber and his classification of types of action and especially the types and forms of rationality. Weber was the most influential exponent of Menger’s ideas, though Weber is almost inevitably taught to sociology students without mention of Menger or any others members of the Austrian school. Alfred Schutz is also taught to sociologists (on account of his subjectivity) without mention of his teacher, Mises.

Weber’s views on rationality can be considered as a refinement of Menger’s list of motivating factors and disturbing influences. Rogers Brubaker wrote a book on Weber “The Limits of Rationality” (Allen & Unwin, 1984) and he identified some sixteen forms of rationality in Weber’s classification.

For example, Weber characterizes various aspects of both ascetic Protestantism and modern capitalism as “rational”. Thus modern capitalism is defined by the rational (deliberate and systematic) pursuit of profit through the rational (systematic and calcuable) organization of formally free labour and through rational (impersonal, purely instrumental) exchange on the market, guided by rational (exact, purely quantitative) accounting procedures and guaranteed by rational (rule governed, predictable) legal and political systems. Ascetic Protestantism is characterized by rational (methodical) self-control and by the rational (purposeful) devotion to rational (sober, scrupulous) economic action as a rational (psychologically efficacious and logically intelligible) means of relieving the intolerable pressures imposed on individuals by the rational (consistent) doctrine of predestination.

Clearly this is an essentialist’s delight, affording material for many theses on the conceptual analysis of the term rationality.

Weber made three distinctions in his study of rationality.

First, a distinction between spontaneous, instinctive, habitual or otherwise unreflective action and deliberate, consciously purposeful action.

Second, between wertrational action and zweckrational action. The former is action “on principle”, oriented to consciously held values. The latter is based on calcuable expectations.

Third, between subjective and objective rationality, the first depending on the perceptions of the actor, the second measuring the effectiveness of the action by some external standard.

In another part of his writing Weber listed four main causes of behaviour.

1. Traditional or habitual action.
2. Affectual, motivated by strong feelings.
3. Wertrational, motivated by a conscious belief in certain values, regardless of the consequences.
4. Zweckrational, motivated by conscious calculations or plans to achieve desired ends with appropriate means.

The point of this diversion is to suggest that if the exegesis of Weber is purged of essentialism his ideas may contribute significantly to the methodology of situatonal analysis.

Koertge and the injection of empirical content into situational analysis.

Birner’s paper provides a summary account of Koertge’s contribution in two papers on this topic. “Popper’s Metaphysical Research Program for the Human Sciences”, Inquiry, 18 (1975) pp 437-62 and “The Methodological Status of Popper’s Rationality Principle”, Theory and Decision, 10 (1979) pp 83-95.

Koertge (1975) wrote “Although Popper is best known for his emphasis on the role of bold empirical conjectures and severe experimental testing in the growth of science, throughout his career he has recognised that bold, but unfalsifiable, metaphysical theories have also been important in the history of science. Atomism, for example, obviously had a profound influence on the growth of science, although according to Popper it only became testable in 1905”.

A note refers to the preface and section 85 of LSD (1959) and various comments on atomism in Conjectures and Refutations. The note also refers to Agassi’s 1964 paper on the roots of scientific problems in metaphysics, and to the way that the demarcation principle loses some of its importance when we realise the large role of non-testable theories.

Koertge went on to consider how we might evaluate metaphysical theories if we cannot test them. She finds three criteria. First, does it solve a problem. Second, does it clash with a well tested scientific theory. Thirdly, evaluation in terms of heuristic power and fertility for science.

Koertge described how situations can be fleshed out by exploring the various factors that contribute to the decisions made by the actors. All the explanatory factors are conjectural but that is no problem for people with a theory of conjectural knowledge. Among those factors are the plans and purposes of the actors, their perceptions of the situation and the means at their disposal. The notion of purpose replaces the assumption of rationality (this is a link with the Austrians). There is no need to assume perfect knowledge, indeed the analysis will often consist of looking for things that the actors did not know about the situation to account for the failure of their plans.

According to the theory of the nature of man which lies behind situational explanations, man’s actions are controlled by his theory of the situation and his decision procedure. Man’s beliefs are controlled by the ideas and information available to him and by his epistemological appraisal procedure. It is because actions and theories are controlled that we may hope to understand them – we may even be able to discover the method in madness…It is because the theories and appraisal procedures are OPEN to correction and improvement that man can become rational in the strong prescriptive sense of the word.

Koertge sums up the paper by suggesting that the model for situational explanations has the same formal (deductive) mode of explanation as the natural sciences and that the heuristic potential of the approach is provided by the metaphysical [and Austrian, praxeological] principle of “man as a rational problem-solving animal”.

Perhaps it would help slightly to refer to man as a “potentially critical and argumentative problem-solving animal” to avoid the multiple ambiguities of the word “rational” and to emphasise the overwhelming importance of language and its higher functions of description and argument to promote the growth of knowledge and the contemplation of alternative futures.

Moving on to Koertge in “The Methodological Status of Popper’s Rationality Principle”, Theory and Decision, 10 (1979) pp 83-95. What does this paper add to the arguments of the previous one?

She examines three explanatory models of human actions, namely Popper, Hempel and Dray. Hempel applied his “covering law” model to human actions.

A was in a situation of type C
A was a rational agent
In a situation of type C, any rational agent will do x

Therefore A did x.

Hempel claims that this is a lawlike situation and the rational agent is to be construed as a “descriptive-psychological concept” which describes certain “broadly dispositional” features of a person. Hempel drew the comparison of the rational agent and the ideal gas.

Dray rejected the covering law model in favour of explanation by “displaying the rationale of what was done”. This looks like an English language version of the continental “verstehan” school of sympathetic or intuitive understanding. These explanations are based on “principles of action” of the form “When in a situation of type C…the thing to do is x”. This has a remarkable resemblance to the general principle that is applied in Hempel’s “covering law” explanation but according to Koertge, Dray made strenuous efforts to distance himself from Hempel.

Dray’s principles can be used to:

licence expectations about an agent’s actions but in his view they do not provide a basis for scientific prediction. They function like rules rather than unbreakable laws and so the principles of action cannot be falsified by finding examples where people failed to act in accord with them. Dray did not accept what he regarded as the scientific usage of causation, rather he accepted Collingwood’s sense “to cause someone to do something is to provide him with a motive for doing it…the necessity of a causal connection, when it is action that we are talking about, is very often rational necessity.

Moving on to Popper, Koertge suggests that it is ironic that Dray considered Popper to be in the scientific/positivist camp with Hempel when he really falls closer to the situational analysis of Collingwood and Dray. This is not irony, it simply reflects the fact that Popper’s views on causation changed over time and the old views (sometimes called the Popper/Hempel model) persist in successive editions of OSE while his more recent views on situational analysis are dogged by contradictory claims and statements about the status of the rationality principle. It is especially confusing when he refers to the rationality principle as an animation principle, analogous to Newton’s Laws in the explanation of celestial motion.

One of the developments that Koertge identified in Popper’s ideas is an increased emphasis on the agent’s perception or theory of their situation. [Koertge does not elaborate on that, but it has at least two aspects. One concerns the values of the actor, exemplified in the difference between the Good Samaritan who crossed the road to help the victim and the Bad Samaritan who crossed the road to see if there was anything left to steal. The other aspect is the appraisal of the situation, what the actor actually picks out as the salient features of the situation, and what was overlooked either through haste or simply because it was out of sight, around the corner or over the hill.]

Koertge suggests the following explanatory schema:

1. Description of the situation. Agent A was in a situation of type C.
2. Analysis of the situation. In a situation of type C the appropriate thing to do is x.
3. Rationality Principle. Agents always act appropriately to their situation.
4. Explanation. Therefore A did x.

Koertge suggests that her reconstruction of Popper’s schema agrees with Dray in recognizing the role of normative appraisals (regarding the appropriate thing to do) however Popper’s scheme has the Hempelian feature of the covering law.

The covering law is the Rationality Principle and it differs from Hempel whose generalization is limited to the particular situation (corresponding to Koertge’s 2, Appraial of the situation). The second difference between Popper and Hempel is that Hempel uses psychological laws for explanation whereas Popper was determined to expel psychologism. [Maybe this was partly motivated by the desire to avoid reduction to theories of psychology that he regarded as highly unsatisfactory and unscientific, also to his desire to avoid subjectivism (which he associated with caprice and arbitrariness). However he could have given more weight to world 2 events because these are real in his three world ontology].

At this point Koertge invokes the three world theory so that “For Popper, the explanation of actions requires a causal interaction between the three worlds”. She then turns back to the content of the rationality principle (which might be better called the principle of purposeful action, or reasoned action [Allen Oakley] or in the language of Talcott Parsons, the voluntarist theory of human action).

Koertge suggests that the Rationality Principle (RP) consists fo two clauses:

i) Every action by a person is a rational [reasoned] response to some problem situation. At this point there is a need for a supplementary theory about detecting problem situations. Then

ii) Every person in a problem situation responds rationally [in a considered way] to it. [Of course this does not preclude ill-considered action which fails to solve the problem or makes it worse. We need to avoid the intrusion of justificationism, in the form of an implicit assumption that the rationality principle has to be a theory of correct or successful action].

For Koertge a rational response [I would prefer “considered”] has three characteristics:

i) arrived at through appraisal of the set of possible solutions
ii) description of the problem and the appraisal process could be verbalised
iii) the person chose as a result of the appraisal procedure (would have taken a better option if it was available).

Koertge notes that there has to be some provision for errors and glitches, both on the part of the actor and in the function of mechanisms and systems that might be activated by the actor. [recall Menger’s list of disturbances including error and laziness].

Consequently Koertge develops the explanation in two stages, first using the Rational Appraisal Principle (the process of deliberation) and then the Rational Action Principle. The scheme is still, to my mind, complicated by using the term “rational” which just carries too much baggage (like “social” in other contexts). We are talking about considered and purposeful action and that opens up the way to find what options were considered, and what purposes the actor had in mind etc.

Koertge then turns to the question of the criticizability of the rationality principle. She quotes Popper “…the attempt to replace the Rationality Principle by another one seems to lead to complete arbitrariness in our model building”. She suggests that the RP may have something like the character of a point of view that is required to write history (on Popper’s account), points of view which cannot be directly tested against evidence, rather they are evaluated in terms of their interest and fertility. There is no doubt in my mind about the fertility of situational analysis because it begs all the questions that need to be asked about events, pointing the way to a pluralistic account of history and society, in contrast with monistic and reductive accounts which attempt to explain everything in terms of some psychological driving force, or a historical theme, or an economic principle.

Koertge reverts to Lakatos language, suggesting that the RP is the hard core of Popper’s progarm. “The positive heuristic is provided by his metaphysical theory of man as an evolving rational problem-solving animal. The program would then be evaluated in terms of its explanatory success in areas such as economics, anthropology and cognitive psychology”. Right on!

That is the approach that characterises the Austrian school of economics, using the principles of praxeology which were described in a previous post. Other people might be able to suggest schools of anthropology and psychology that take this approach.

This was the approach that Talcott Parsons introduced into sociology in his first book, “The Structure of Social Action” (1937) though his work did not develop especially well after that, for reasons which I will try to explain.

Rafe Champion 2003


Written by Admin

November 15, 2005 at 7:15 pm

Posted in Uncategorized

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: