Copyright © 1999 Massachusetts Institute of Technology
Functional decomposition is the analysis of the activity of a system as the product of a set of subordinate functions performed by independent subsystems, each with its own characteristic domain of application. It assumes that there are a variety of functionally independent units, with intrinsically determined functions, that are minimally interactive. Functional decomposition plays important roles in engineering, physiology, biology, and in artificial intelligence. Functional morphologists, for example, distinguish the causal or functional roles of structures within organisms, the extent to which one structure may be altered without changing overall function, and the effects of these structures for evolutionary change. Within cognitive science, the assumption is that there are a variety of mechanisms underlying our mental life, which are domain specific and functionally independent. The classical distinction in DESCARTES between understanding, imagination and will is a functional decomposition, which postulates at least three independent faculties responsible for specific mental functions; likewise, the distinction drawn by KANT between sensation, judgment, understanding, and reason offers a partitioning of our mental faculties based on their cognitive functions, and is equally one that postulates a variety of independent faculties responsible for specific mental operations. In more recent psychological work, the distinction between sensory stores, short-term MEMORY, and long-term memory elaborated by Richard Atkinson and Richard Shiffrin (1968) is a functional decomposition of memory, based on their domains of application (for a classic source, see Neisser 1967).
Functional decomposition typically assumes a hierarchical organization, though a hierarchical organization is consistent with different modes of organization. Thus, the mind is conceived as having a modular organization (cf. MODULARITY OF MIND), with a variety of faculties, each with independent, intrinsically determined functions. Each of those modules in turn may have a modular organization, with a variety of independent, intrinsically determined functions. Sensory systems are relatively independent of one another, and independent of memory, language, and cognition. Language in turn may be taken to consist of a variety of relatively independent subsystems (cf. MODULARITY AND LANGUAGE), including modules responsible for PHONOLOGY, PHONETICS, SEMANTICS, and SYNTAX. The extent to which a hierarchical organization, or functional independence, is realistic can be decided only empirically, by seeing the extent to which we can approximate or explain system behavior by assuming it.
Functional decomposition is easily illustrated by appealing to the understanding of language. In the early nineteenth century, Franz Joseph Gall (1758-1828) defended the view that the mind consists of a variety of "organs" or "centers," each subserving specific intellectual or moral (that is, practical) functions, with dedicated locations in the cerebral hemispheres. These intellectual and moral functions were sharply distinguished at a higher level from the "vital" functions and affections that Gall located in the "lower" portions of the brain, and the specific functions in turn were distinguished from one another. There were differences between Gall and his fellow phrenologists concerning the number and characterization of the specific faculties, but within the intellectual faculties, phrenologists typically distinguished broadly between the external senses, various "perceptive" faculties (including faculties for perceiving weight, color, tune, and language, among others), and the "reflective" faculties constitutive of reason. The primary faculties were the species of intellection, and they were assumed to belong to specific organs in the brain. Gall held an extreme view, assuming that the basic functions were strictly limited in application, invariable in their operation, and wholly independent of the activities of other faculties. That is, he assumed that the mind was simply an aggregate of its independent functions and that there was no overlap or interaction between these organs. Because he recognized no interaction between the faculties, complex abilities became simply the aggregates of simple abilities. Gall assumed, in other words, that the mind was both hierarchical and aggregative, or simply decomposable.
Paul Pierre BROCA (1824-1880) was also a defender of "organology," retaining both the discrete localizations of the phrenologists and the view that the size of organs was responsible for differing mental abilities. Following Jean Baptiste Bouillard (1796 - 1881), Broca emphasized the importance of dysfunction in determining functional organization. By August 1861, Broca had described in some detail the anatomical changes accompanying a disorder of speech that he called "aphemia," and that we would describe as an APHASIA. The patient, known as "Tan," lost the ability to speak by the time he was thirty, and over the years his case degenerated. Broca relied on interviews with the hospital staff to discover that Tan's initial "loss of articulate language" was due to a focal lesion in the frontal lobe. Broca's conclusion was that there were a variety of "organs" corresponding to discrete mental functions. Karl Wernicke (1848 - 1905) subsequently elaborated the basic model, reframing it in terms of an associationistic psychology rather than a faculty psychology and distinguishing sensory and motor aphasias. Wernicke concluded that there was a series of discrete loci mediating the comprehension and production of speech. On the basis of clinical observations, Wernicke concluded there were three distinctive "centers" associated with language use: a center for the acoustic representations of speech, a center for motor representations of speech, and a center for concepts typically mediating between the two. Disruptions of the various associations between these centers resulted in the various aphasias. The resulting functional decomposition for language use thus had at least three components, and a linear organization: the output of one "organ" serves as the input for the next, though the function performed or realized by each module is intrinsically determined. This basic model has since been elaborated by a number of clinical neurologists, including Norman GESCHWIND. The organization is no longer aggregative, but sequential, with relatively independent functional units. This is near decomposability.
A commitment to functional decomposition has continued in a variety of forms in more recent work in cognitive science, including the new "organology" of Noam Chomsky (1980), the "modularity" defended by Jerry Fodor (1983), and the FUNCTIONALISM of William Lycan (1987). Steven Pinker (1994), for example, argues that language is an ability that is relatively independent of cognitive abilities in general. The clear implication of such independence is that it should be possible to disrupt linguistic abilities without impairing cognition, and vice versa. The studies of aphasia exhibit such patterns. A commitment to some form of functional decomposition or modularity might seem inevitable when dealing with a phenomenon as complex as mental life. Herbert Simon (1969) has emphasized the importance of simple decomposability and near decomposability, as well as of hierarchical organization in complex systems. In explaining the behavior of a complex system, it is often possible to establish independent functional characterizations for components, ignoring both the contributions of other components at the same level as well as the influences operative at higher or lower levels. This is, however, not always true, and the cases are often more complex than they might initially appear.
Numerous examples of functional decomposition are available from recent work in cognitive science. It is common, as noted, to analyze memory into distinctive subsystems. Commonly, it is assumed that there are at least two stages, presumably with discrete physiological mechanisms: the first process is short-lived, lasting from minutes to hours, and the second is of indefinite duration. Conventionally, this distinction between short-term and long-term memory is assayed by recall tests. This is by no means the only decomposition of memory, and is anything but unproblematic; more specifically, the experimental evidence leaves it unclear whether the distinction between short- and long-term memory is a distinction between modules, or modes of processing, and whether short-term memory is a unitary entity. Experimentation in memory typically involves some measure of retention based on recall or recognition of some predetermined material, and more recently using dual tasks in parallel. This research has led to a variety of ways of understanding the organization of memory, including distinctions between working memory, semantic memory, and declarative memory. There is currently no clear consensus concerning the most appropriate theory, and no model that naturally accommodates the entire range of the phenomena. In a similar way, linguistic competence is generally understood as the product of a set of distinct subsystems. Wernicke's distinction between comprehension and production has been replaced with distinct processes involved in language use, typically distinguishing between semantic and syntactic functions. Again, this decomposition is not unproblematic, and there is some evidence suggesting that such decompositions do not yield functionally independent subsystems.
Atkinson, R., and R. Shiffrin. (1968). Human memory: A proposed system and its control processes. In K. W. Spence and J. T. Spence, Eds., The Psychology of Learning and Motivation, vol. 2. New York: Academic Press.
Broca, P. (1861a). Perte de la parole. Bulletins de la Société Anthropologie 2:235-238.
Broca, P. (1861b). Remarques sur le siêge de la faculté suivies d'une observation d'aphémie. Bulletin de la Société Anatomique de Paris 6:343-357.
Chomsky, N. (1980). Rules and Representations. New York: Columbia University Press.
Fodor, J. A. (1983). Modularity of Mind. Cambridge, MA: MIT Press/Bradford Books.
Lycan, W. (1987). Consciousness. Cambridge, MA: MIT Press/Bradford Books.
Neisser, U. (1967). Cognitive Psychology. New York: Appleton-Century-Crofts.
Pinker, S. (1974). The Language Instinct. New York: William Morrow.
Simon, H. A. (1969). The Sciences of the Artificial. Cambridge, MA: MIT Press.
Wernicke, C. (1874). Der Aphasiche Symptomcomplex: Eine Psychologische Studie auf Anatomischer Basis. Breslau: Cohen and Weigert.
Amundson, R., and G. V. Lauder. (1994). Function without purpose: the uses of causal role function in evolutionary biology. Biology and Philosophy 9:443-470.
Bechtel, W., and R. C. Richardson. (1993). Discovering Complexity. Princeton: Princeton University Press.
Bradley, D. C., M. F. Garrett, and E. Zurif. (1980). Syntactic deficits in Broca's aphasia. In D. Caplan, Ed., Biological Studies of Mental Processes. Cambridge, MA: MIT Press, pp. 269-286.
Cummins, R. (1983). The Nature of Psychological Explanation. Cambridge, MA: MIT Press.
Gregory, R. L. (1961). The brain as an engineering problem. In W. H. Thorpe and O. L. Zangwill, Eds., Current Problems in Animal Behaviour. Cambridge: Cambridge University Press, pp. 307-330.
Gregory, R. L. (1968). Models and the localization of function in the central nervous system. In C. R. Evans and A. D. J. Robertson, Eds., Key Papers: Cybernetics. London: Butterworths.
Gregory, R. L. (1981). Mind in Science. Cambridge: Cambridge University Press.
Johnson-Laird, P. N. (1983). Mental Models. Cambridge: Harvard University Press.
Schacter, D. L. (1993). Memory. In M. I. Posner, Ed., Foundations of Cognitive Science. Cambridge, MA: MIT Press, pp. 683-725.
Copyright © 1999 Massachusetts Institute of Technology