Explicating and Illustrating The Distinction

The debate about the virtues and vices of functional approaches to behavioural science on the one hand
and cognitive‐process approaches on the other hand is as old as the cognitive revolution against
Skinnerian behaviourism or Paul Meehl's critique of Brunswik's probabilistic functionalism. The tone of
this debate was often polemic and lacking respect. Each of the two camps seemed to presuppose that
the other camp misses even the most important insights. For instance, Meehl's (1961) famous notion
that correlation is an instrument of the devil means, basically, that Brunswik's (1955) lens model is not
just worthless but, due to its reliance on environmental correlations, even of negative value. Conversely,
when Skinner (1985) wrote “I accuse cognitive scientists of misusing the metaphor of storage …
speculating of internal processes … relaxing standards of definition and logical thinking …,” drawing on
Émile Zola's famous exclamation “J'accuse …” in the Dreyfus Affair , he meant that cognitive psychology
is causing damage to the field. However, this polemic style should not prevent one from reflecting on
the distinction between functional and cognitive psychology (cf. De Houwer, 2011; De Houwer, Barnes‐
Holmes, & Moors, 2013; Fiedler, 2014).
An analysis of the state of the art in modern psychological science in 2014 reveals that there is still a
good deal of conflict between the two camps—functional psychology and cognitive psychology—and
that there is still no common ground for an open‐minded methodological and meta‐theoretical
discussion of how both approaches might complement and fertilise each other. In the present article, I
will start with a brief outline of functional and cognitive psychology and of what I consider to be their
major themes. In doing so, I will not refrain from a comparative evaluation. To anticipate, my evaluation
will not be a shady compromise saying that both approaches have their virtues and their vices. Rather, I
will honestly express my disappointment with the paucity of cogent evidence that cognitive psychology
has found out about the properties of the behaviourists' black box. At the same time, I will clearly state
that when real progress has been attained it has been mainly progress at the functional level of relating
behavioural phenomena to environmental causes and boundary conditions. Functional rules are logically
antecedent and superior to cognitive‐process rules, which are always contingent on a functional task
setting. Whenever some clever cognitive model or research design result in strong evidence, validation
studies have to return immediately to the functional level. But despite its apparent primacy, the
functional approach is itself dependent of cognitive research, which is sorely needed to understand the
psychological meaning and validity of functional relationships.
EXPLICATING AND ILLUSTRATING THE DISTINCTION
Let us illustrate the characteristic features of functional and cognitive psychology with a few concrete
examples. The functional research program is deeply anchored in the behaviourist tradition. The goal is
to analyse manifest behavioural outputs measured in the dependent variable as a function of
environmental stimuli and experimental manipulations specified in the independent variable. For
example, in the prototypical case of psychophysics (cf. top row of Table 1), the goal is to specify relations
between subjective quantities (life satisfaction, subjective distance, utilities) and their physical
analogues (income, objective distance, monetary value). Such research is by no means confined to
assessing simple relations between loudness and sound pressure or between subjective and objective
stimulus durations. Psychophysical functions can be highly sophisticated and multi‐factorial, describing
the relativity of responses to particular stimuli in the context of other stimuli, subjective expectations or
dimensional constraints imposed by response scales (Garner, 1976; Mellers & Cooke, 1994;
Parducci, 1965). For instance, in Mellers, Schwartz, and Ritov's (1999) decision‐affect model, the
subjective attractiveness of decision options depends not only on the (monetary) gains or losses
obtained with the chosen option but also on the range and the distribution of the alternative options'
outcomes, thus giving the psychophysical model a feature of relativity. In the Garner (1976) paradigm,
another example of refined psychophysics, the performance on a speeded classification task depends
not only on the discriminability of the stimuli on the task‐relevant dimension but also on the redundancy
with other attribute dimensions and on the integral versus separable relationship that exists between
dimensions. In any case, though, the functional approach is obliged to relating measurable dependent
variables to interventions and controllable manipulations of independent variables.

Explicating and Illustrating The Distinction

Uploaded by

Copyright:

Available Formats

Explicating and Illustrating The Distinction

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Explicating and Illustrating The Distinction

Uploaded by

Copyright:

Available Formats

The debate about the virtues and vices of functional approaches to behavioural science on the one hand

EXPLICATING AND ILLUSTRATING THE DISTINCTION

You might also like