[Note for bibliographic reference: Melberg, Hans O. (1994), Two approaches to
explanation: Verstehen or Correlation?, http://www.oocities.org/hmelberg/papers/940423.htm]
[Comment: This is the first serious paper I wrote on methodology. My views have developed
greatly since I wrote this paper, as have (hopefully) the quality of my arguments, but I
am still critical of the functionalist approach.]
Two approaches to explanation:
Verstehen or Correlation?
Interpretive-individual or Structural-functional explanation
One of the central debates in the philosophy of the social sciences is whether one should
explain social phenomena by reference to structural or individual factors. This essay
argues that the best approach to this issue is to view the individual and the structural
approach as complementary. This, of course, is no revolutionizing argument. For example,
Anthony Giddens has proposed a theory of structuration in which he claims to unite both
individual and structural factors. Thus, in an attempt to go a bit further than the rather
bland "both are important" argument, this essay will try to examine the factors
that determine to which extent explanation should be coached in individualistic or
structural terms. The essay will present these arguments mainly through a critical
examination of Jon Elster's views on functionalism.
The structural-functionalist argument
What is the structural-functional approach and what are the circumstances in which we
consider a structural functional approach to explanation to be valid? Consider the
following example from biology:
Phenomenon: A decline in the population of bright butterflies and an increase in the
population of dark butterflies between time t and t+k within a limited geographic area.
Explanation: At time t a new factory started production in the area. Because of the smoke
pollution from this factory the physical environment in the area became darker.
Consequently, the bright butterflies became more visible and hence the birds could catch
the bright butterflies more easily than before while it became more difficult to catch the
dark butterflies. This led to a decrease in the population of bright butterflies and an
increase in the population of dark butterflies.
The above example illustrates the meaning of structural-functionalist explanation: A
change in the environment led to a change in the composition of butterflies. The more
"functional" survived, the less functional did not. By the term structure it is
here meant those conditions external to the individual. Functional means beneficial with
respect to some goal such as survival. If this type of explanation is acceptable in
biology, what makes it invalid for explaining social phenomena?
Jon Elster claims that two factors make the functional approach invalid for the social
sciences. First, "selection processes [in social science] work too slowly to produce
behaviour that is optimally adapted to a rapidly changing environment." Second, human
action may intentionally change the environment, thus explanation should not rest upon the
analogy that we are purely passive victims of structural change.1 I
will deal with these in more detail.
Fast changing environments relative to selection mechanism
a) Are social environments really fast changing?
One objection to Elster is that all social environments are not fast changing. However,
judging by the number of footnotes Elster is not an author who throws in an argument
without reflecting on it. Accordingly we should examine what might lie behind his claim.
Consider the following:
There are two types of agents (P1 and P2) with different sets of characteristics (A-F):
P1: A, C, D
P2: A, B, E, F
A change in the environment may make characteristic D more functional ie. having D is in
some way beneficial in the new environment. Furthermore, there might be another structural
change making B more functional and C dysfunctional. Does this mean that we are unlikely
to see a clear change in the composition of P1 v. P2 in the environment?
The answer depends on the degree of the shift ie. we need to know how much more
"functional" P1 becomes as a result of having B and how much less functional it
is to have characteristic C in the new environment. Hence, it is not only the speed of
environmental change that matters, but also the quantity of change. As an illustration,
consider Parsons claim that the industrial revolution changed the structure of the family
from extended to nuclear because a nuclear family was more functional in an industrial
society. It would be incorrect to argue that his argument is invalid just because there
were many other environmental changes going on at the same time (i.e. that the environment
was fast changing). First, this would require the other changes to work in the opposite
direction (i.e. in favour of the extended family). Second, it would require that the
quantitative effect of these changes outweighed the quantitative effects of the industrial
revolution. Hence, speed matters but so does quantity and direction.
b) Why speed might not destroy the selection mechanism: Flexibility
To argue that some environments change so fast that the social selection mechanism does
not have time to "eliminate" the dysfunctional agents, overlooks the possibility
that speed itself may make some individuals functional or dysfunctional. Some individuals
have more of those characteristics that are functional in rapidly changing environments.
For example, ability to learn quickly might lead to greater flexibility in the face of
speedy environmental change. Rationality, intelligence, creativity, diversity of skills
and even religion are other examples of characteristics that might be functional in fast
changing environments. It is important to note here that they are functional precisely
because the environment is changing. This implies that agents with less of these skills
are less functional in fast changing environments. Thus, speed in itself does not make
functional explanations invalid.
c) How fast is fast?
In an effort to determine when fast environmental change relative to the selection
mechanism would make functional explanations inappropriate, I wrote a computer program.
Assume: that there are two types of agents (P1 and P2); there is a fixed initial number of
type 1 and type 2 agents alive; That P1 is functional in environment 1, while P2 is
functional in environment 2; that the selection mechanism reduces the population which is
dysfunctional by x% while increasing the functional population by x% between each time
period; That the speed of the environmental change between environment 1 and 2 is 0-100
where 100 is a change every time period, while 0 is no change ever. Given these
assumptions I first explored the consequences of unconscious actors (in Program A), and
second of conscious actors capable of learning (Program B). These programs do not in any
way "prove" any conclusions, however they are useful as a heuristic device.
Unfortunately (and maybe not too surprisingly) I was unable to draw any reliable and
sensible conclusions from these experiments. At a theoretical level is seems clear that as
long as the speed of the selection mechanism is higher than the speed of the relevant
environmental change, functional arguments are still valid.
d) The selection mechanism (and its speed)
In the example of the butterflies, the dysfunctional (bright) butterflies were
"eliminated" by the birds. In the social world people do not die because they
are dysfunctional (unless in extreme cases). Neither do dysfunctionality impair people's
reproductive capacity. In fact one might argue that the opposite is the case in the social
world. Social security could by a cynic be viewed as rewards to dysfunctionality and
having low income, a possible sign of less functional ability, is often correlated with
having many children. Hence, there is a problem in that functional explanation require
some kind of social selection mechanism, while this is difficult to find in practice.
It should be admitted that there is at best an imperfect social selection mechanism
analogous to the natural selection mechanism. Bankruptcy is one such imperfect social
selection mechanism. War may also be viewed as a selection mechanism. More generally,
economic constraints make sure that certain environments are populated by people who are
"functional" in that environment: Largely irrational people do not survive for
long as stock brokers; Boring comedians do not get the audiences needed to survive as
comedians; Seriously inefficient firms do not survive competition from other firms.
However, these constraints are only imperfect because the mechanism for selection might be
slow (as Elster suggests) and it may require a large quantity of dysfunctionality to be
eliminated. Hence, this imperfect social selection mechanism only establishes a lower
bound on possible outcomes of social processes.
Bringing the individual back in
A more promising avenue than to argue for social selection by elimination, would be to
argue for "selection" by signalling. In other words there is only a very weak
social selection mechanism that forces "exit", but there is a "voice"
signal.2 For example, being dysfunctional might in some environments
mean lower income or more discomfort (eg. isolation, sanctions). This provides a
justification for bringing the individual back into the explanation because when the
creative and intentional individual observes that income is falling or an increase in
social problems (rows, friends leaving etc.) he/she would try to adapt.3
The functionalist perspective did not have any room for this because adaptation was
ensured through external elimination of passive, unchanging, and dysfunctional agents.
Viewing the structure as shaping signals in an environment of intentional and creative
individuals opens up the possibility of a synthesis of structural and individual
approaches to explanations.
The synthesis
Based on the arguments so far the synthesis between individual and structural explanation
can be stated as follows. The profitability of various actions is largely determined by
the structure. However, this does not mean that individual actions are completely
determined by the structure (or fully explicable through structural factors). Since the
individual is not forced by the selection mechanism to choose the functional alternative,
the intentional and creative individual may always choose the dysfunctional option.
However, this is a weak argument against structural explanations because one might
question the actual opportunity of the individual to choose the dysfunctional alternative.
Imagine the following situation determined by the structure:
Act Consequence
A Receive £100
B Being shot
If this was the situation determined by the structure, there is no need to bring
intentionality back in. However, as argued, the structure is not as restrictive as in the
choice between £100 and death. Rather it was a choice between more or less outcome
(signals, not selection):
Act Consequence
A Receive £100
B Pay £100
However, there is still little room for intentionality in this explanation because the
alternatives A and B are not worth the same. As Joseph Raz has argued, a choice is not
meaningful unless it is assumed that there are several valuable alternatives to choose
between.
Hence, for the individual (intentionality, belief, meaning) to play an important role the
structure must leave at least two valuable and feasible alternatives. This, of course, is
more possible in some environments than others. For example, environments in which
outcomes are incommensurable (or close to) leave more scope for intentionality than
environments in which outcomes are measured on a single uniform scale (monetary price) and
where preferences are identical.
The structure is also less important as an explanatory factor when individuals are
uncertain about the true structure and hence have to rely on beliefs. Thus, the obvious
conclusion that the less we know about the environment, the more important beliefs and
intentions are in an explanation of why a particular act was chosen.
Another rather obvious conclusion is that the structure is less important the larger
influence the single individual has on the structure itself.
Implication
Correlation and Verstehen and reliable routes to knowledge
The theoretical validity of structural-functional approach is only one of several criteria
for evaluating this type of explanation. Another criterion is reliability ie. the
probability that the explanation leads to true knowledge. This is important because
functional explanations, as presented by G. A. Cohen, uses correlation to establish that A
is functional to B (ie. it is argued that A is functional to B if whenever A would be
functional to B then A appears. The way to discover this would be by correlation analysis
of historical facts - verbally or statistically)4 . Elster, on the
other hand, demands the mechanism that would make it clear exactly how A was functional to
B, a demand which would require attention to individual intention and beliefs. This essay
has argued that both approaches are valid, but to various degrees depending on the context
of the phenomena to be explained. However, one should note the serious problem of spurious
correlation5 which makes correlation analysis less reliable. As a
route to knowledge Weber's concept of Verstehen (i.e. understanding of how a phenomena
makes sense from individual's point of view) is arguably more reliable. However, the point
is that correlation of aggregates are complementary routes to knowledge about social
phenomena. Ideally, of course, both should be provided. As an inaccurate rule of thumb one
might say that the Verstehen approach suggest the reliability of knowledge, while the
correlation approach suggest the general applicability of the explanation (knowledge).
Though, this is more a slogan than a rule.
Conclusion
Elster claims that there are three categories of explanation: Causal, Intentional and
Functional. This essay suggests that at least the intentional and the functional approach
are not mutually exclusive, and that a good explanation should pay more attention to one
over the other depending on the nature of the phenomenon. Finally, the issue of
reliability of knowledge indicated that there might be good reasons to put more emphasis
on the individualistic approach and less on Cohen's structural/correlation approach.
NOTES
1. "...much sociobiological though neglects a massively important feature of human
beings: their creativity or capacity for problem solving." p. 78 (Nuts and Bolts)
2. Hirschman's terminology fits strikingly, but not perfectly since the structure does not
actually use "voice" in Hirschman's sense (intentionally).
3. In the same way that the social selection mechanism is not perfect, the signalling
mechanism is not perfect because individuals may draw the wrong lesson from a signal in
complex environments. Thus, the quality of learning processes becomes important for the
well functioning of social signals.
4. This is C. A. Cohen's "Consequence Law." (From A. Callinicos (ed.)
"Marxist Theory", p. 98)
5. Spurious correlation exists when two series are correlated without being causally
connected. This is very common, hence it is very difficult to use correlation as a basis
for judging which factors are functional.
Bibliography
Boudon Raymond (translated by J. C. Whitehouse), Theories of Social Change, Polity
Press, Cambridge, 1986 (French 1984)
Callinicos A. (ed.), Marxist theory, Oxford University Press, Oxford, 1989
Cohen, G. A., Karl Marx's theory of history: A Defense, Oxford University Press,
Oxford, 1978
Durkheim Emilie, The Rules of Sociological Method, Free Press, Glencoe, 1938
(original 1893)
Durkheim Emilie, Suicide, Routledge & Kegan, London, 1952 (original: 1897)
Elster Jon, Making Sense of Marx, Cambridge University press, Cambridge, 1985
Elster Jon (ed.), Rational Choice, Cambridge University press, Cambridge, 1986
Elster Jon (ed), Nuts and Bolts for the Social Sciences, Cambridge University
Press, Cambridge, 1989
Elster Jon, The Cement of Society, Cambridge University Press, Cambridge, 1989
Hirschman Albert O., Exit, Voice and Loyalty, Harvard University press, Cambridge
Massachusetts, 1970
Giddens Anthony, Studies in Social and Political Theory, Hutchinson, London, 1977
Giddens Anthony, Sociology, Polity Press, Oxford, 1993
Hausman Daniel, The Inexact and Separate Science of Economics, Cambridge University
Press, Cambridge, 1992
Merton Robert, Social Theory and Social Structure, Free Press, Glencoe, 1957
Parsons Talcott (ed), Theories of Society: Foundations of Modern Sociological Theory,
Free Press, 1987
Raz Joseph, The Morality of Freedom, Oxford University Press, Oxford, 1986
Ryan Alan (ed.), The Philosophy of Social Explanation, Oxford University press,
1973
Weber Max, The Protestant Ethic and the Spirit of Capitalism, Allen and Unwin, 1920
(German: 1904)
[Note for bibliographic reference: Melberg, Hans O. (1994), Two approaches to explanation:
Verstehen or Correlation?, http://www.oocities.org/hmelberg/papers/940423.htm]