How do representations gain meaning?

Daniel Stern postulated six internal formats that pick up various aspects of the incoming sense data. Probably not aware of it, Stern repeats Aristotle’s problem of how the pieces from different sensory channels are parsed together. By being so distinct and clear, Stern finds exactly the right words to show the problem. I must repeat the passage from my previous post:

...the diverse events and feelings are tied together as necessary elements of a single unified happening that, at one of its higher levels, assumes a meaning. The problem with this and other such solutions is how the meaning, even a very primitive one, slips in or gets assigned or is constructed from the pieces. (Stern: The motherhood constellation, p. 89, my italics)

Indeed, by what mental function does meaning become attached to the representation, or the internal perceptual form of an object or an event? This is a typically Aristotelian problem, as Kaukua and Kukkonen pointed out in their article. An interesting Medieval attempt to solve it was presented by Ibn Sina (the Latin Avicenna), a Persian physician who lived around the turn of the first millennium (980–1037). He was an important developer of Arabian philosophical psychology and a link in the subsequent Western Medieval adoption of Aristotle’s philosophy.

Much in the same manner as Stern, Ibn Sina recognised the need to postulate a number of internal functions, or internal senses, that processed the elemental data, received by the five sense organs. I managed to retrieve a translated excerpt  in which Ibn Sina enlists the five internal senses that perform various synthetic operations on the incoming forms.

One of the animal internal faculties of perception is the faculty of fantasy, i.e., sensus communis, located in the forepart of the front ventricle of the brain. It receives all the forms which are imprinted on the five [external] senses and transmitted to it from them. Next is the faculty of representation located at the rear part of the front ventricle of the brain, which preserves what the sensus communis has received from the five senses even in the absence of the sensed object. … Next is the faculty of the 'sensitive imagination' in relation to the animal soul, and the 'rational imagination' in relation to the human soul. This faculty is located in the middle ventricle of the brain near the vermiform process, and its function is to combine certain things with others in the faculty of representation, and to separate some things from others as it chooses. Then there is the estimative faculty located in the far end of the middle ventricle of the brain, which perceives the non-sensible intentions that exist in the individual sensible objects, like the faculty that judges that the wolf is to be avoided and the child is to be loved. Next there is the retentive and recollective faculty located in the rear ventricle of the brain, which retains what the estimative faculty perceives of the non-sensible intentions existing in individual sensible objects. (Avicenna, translated in Rahman, F.(1952). Avicenna's psychology. London: Oxford University Press,  p. 31)

Ibn Sina is surprisingly modern by regarding these internal faculties as brain modules. Certainly, current brain imaging studies have shown that these modules  of the “sensible soul” do not reside in the ventricles, but the contemporary modular theories of mind follow exactly the same logic of analysis as Ibn Sina does in the excerpt. Going back to Stern’s question about constructing meaning of events I want you to pay attention to the estimative faculty that is capable of perceiving the non-sensible intentions that exist in sensible objects.

Ibn Sina understood well the impoverished character of the Aristotelian sensible forms. He asked, why does the lamb flee when perceiving a wolf. The grey, hairy form does not “inform” a particular danger. Something had to be added, and here we have a conception of meaning that “exists in the individual sensible objects” and is recognised by the estimative faculty (cf. Stern’s “protonarrative envelope”).  Thus, in addition to forms, objects seem to “emit” intentions that are perceived by an internal mental faculty. Actually, the currently popular research on the “theory-of-mind”, or "mind-reading" module represents a truly Avicennian approach to understanding how we make sense of others. I will return to this topic in due course.

According to Kaukua and Kukkonen, Ibn Sina had a sophisticated understanding of intentions.  He did not, actually, regard them as properties of objects or events that they "emit" and are then sensed by the estimative faculty. Instead, intentions vary according to the reciprocal positioning of the agent an object.  Kaukua and Kukkonen illustrate this by an example of the hunter’s way of perceiving a wolf. While the lamb flees, the hunter has a rifle and this alters the intentional balance. Now it is rather the wolf who must escape. But the hunter may be an environmentalist and does not want to harm the wolf. “Acts of meaning” (in Bruner’s terms) are not anything as such. They receive their meaning only in the concrete relationships that are established between the actor and the object.

Representing interaction

In Chapter 5 of The Motherhood Constellation, Daniel Stern traces the path from mother’s overt behaviour to the infant’s subjective experience, or his way of representing  interpersonal events. Stern is aware of the difficulties ahead: “When thinking about infants, we are far less sure what such representations are and how they get there" (i.e., in the infant’s mind). (p. 79). The first problem is the nature of social interaction. Inanimate events can be described in the standard Aristotelian model of incoming forms:

There are several important differences between the representations of inanimate physical happenings and those of subjective interpersonal happenings. For inanimate physical happenings, the mental events in a representation are thought to be isomorphic with the real events: they are simply performed virtually on an internal stage. (Stern, The motherhood constellation, 1995, p. 80)

Joint interpersonal actions have several features that make this straightforward model inadequate. Stern lists six, which I will not reproduce here. Taken together, they make up a particular format of, or a placeholder for representations that Stern calls “schema-of-being-with-another”. It can be seen as an elaboration of his earlier description of “Representations of Interactions that have been Generalized” or RIGs.  Stern warns of equating representations with the relatively simple idea of internalising interactions and adopts a strictly Kantian position:

...these representations are not formed from external events or persons that have  been internalized. They are not put inside from the outside. They are constructed from the inside, from the self-experience of being with another. Nothing is taken in. (ibid., p. 81)

I admire Stern’s bold precision even if by being so explicit he paints himself into the corner. Now he is faced with the issue by what means does the infant construct the internal representations of interactive events. In a truly Kantian manner Stern has to assume a number of innate "principles of judgment” or schemas that do the job for the infant.

What fundamental formats for representing already exist to account for each of the basic elements of an experience as well as to tie them together? Taking each element separately, we have perceptual schemas (e.g., visual images) and we have conceptual schemas (e.g., symbols and words)... Piaget introduced the sensorimotor schema, which added motor acts and their coordination with sensory experiences as yet another fundamental way of representing experience. More recently, another basic form of human representation available to children has been added, namely, an invariant sequence of events that is represented as a single script...  (ibid., p. 82)

These four representational formats are necessary but not sufficient for the construction of the schema-of-being-with-another. Hence, Stern adds two more. The first is a placeholder for the temporal contour of affects, called “feeling shape” and the second has the intriguing name of a “protonarrative envelope”. It is needed in order to make sense of the five other formats in terms of meaning:

...the diverse events and feelings are tied together as necessary elements of a single unified happening that, at one of its higher levels, assumes a meaning.

The problem with this and other such solutions is how the meaning, even a very primitive one, slips in or gets assigned or is constructed from the pieces. There is a way out of this dilemma, and that is to assume that there is yet another fundamental way of representing human events, a sixth schema made up of “acts of meaning”, as Bruner and others have argued. (Ibid., p. 89)

As Jari Kaukua and Taneli Kukkonen pointed out in their article ”Sense perception and self-awareness” , an Aristotelian model of sense perception unavoidably necessitates an assumption of mental faculties that unite the disparate information conveyed by the senses. “What is it that allows us to go beyond mere sensations of red, fragrant, and smooth and to recognise an apple for an apple?” (Kaukua & Kukkonen, 2007, p. 96). Stern’s list of six representational formats summarises modern conceptions in cognitive developmental psychology, but as a solution to the puzzle it has ancient roots. I will get back to these in my next post.

John Locke's Aristotelian rift

Daniel Stern is certainly not the first to raise the question of how mental representations are expressed. Contemporary philosophers have a convenient way of dealing with Aristotle’s rift between incoming forms and their signification by words. They speak about epistemic and semantic aspects in the mind’s operations. Separating these aspects allows them to bypass the problem how mental representations are transformed into words.

This is not, however, a novel solution. I recognise it as an "analytic jump" in John Locke’s An essay concerning human understanding between the Books II and III. Book II deals with the path from sense operations to the formation of mental ideas:

The soul begins to have ideas when it begins to perceive. To ask, at what time a man has first any ideas, is to ask, when he begins to perceive; - having ideas, and perception, being the same thing. (Locke, Book II, Chapter 1. Of ideas in general and their origin). 

Locke’s analysis of how ideas are generated in Book II soon becomes quite complex. Not all ideas stem from “privative causes”. Bodies have primary qualities, secondary qualities, etc. This complexity does not, however, alter the fact that the analysis begins from sense perception, which is largely understood in an Aristotelian way as a passively receptive process. Then, the very first page in Book III introduces human expression by the following quote:

 Man fitted to form articulate sounds. God, having designed man for a sociable creature, made him not only with an inclination, and under a necessity to have fellowship with those of his own kind, but furnished him also with language, which was to be the great instrument and common tie of society. Man, therefore, had by nature his organs so fashioned, as to be fit to frame articulate sounds, which we call words.

To use these sounds as signs of ideas. Besides articulate sounds, therefore, it was further necessary that he should be able to use these sounds as signs of internal conceptions; and to make them stand as marks for the ideas within his own mind, whereby they might be made known to others, and the thoughts of men’s minds be conveyed from one to another.”

 

Locke dodges the problem of how representational ideas transform into expressive signs by referring to God’s kindness. As modern secular scientists, psychologists cannot solve the problem in this convenient way. Hence, Stern is quite right when he points out in “The motherhood constellation” that the transformation of mental representations into expressions remains an enigma.  

Behavior or expression?

The quoted passage from Daniel Stern’s book contains so many issues that several posts may be needed. I will begin with the first transformation in “chasing the path” from maternal representations to the infant’s understanding of its intent. Here it is: “We must be able to describe in concrete behavioral terms how the mother manages to be rejecting and aloof such that her rejection and aloofness can be perceived by the baby and have an impact on him.”

How do psychic processes become external? The concepts by which we attempt to describe this event affect the outcome. If we carry out the analysis by using “representation” and “behavior” we will get into Stern’s question, and the puzzling problem in it. In fact, the very term transformation contains an Aristotelian reverberation. A form is transferred from one domain into another.

An alternative would be to regard “the transformation” as an expression. Psychic processes become manifested through expressions. I Consulted the Oxford English Dictionary to get the current meanings of the term and found the following:

The action of expressing or representing (a meaning, thought, state of things) in words or symbols; the utterance (of feelings, intentions, etc.).

The action or process of manifesting (qualities or feelings) by action, appearance or other evidences or tokens.

An action, state, or fact whereby some quality, feeling, etc., is manifested; a sign, token

The confusion between representation and reference seems to be so pervasive that it is even reproduced in the OED definitions. “Expressing or representing” are used as equivalent terms. This won’t do! There is an unrecognized internal ambiguity here that reproduces Aristotle’s original rift between in-dwelling forms and their expressions in words. How does a form become a sign within the mental apparatus of a human being?

The OED definition clearly wants to include signs in characterizing expressions. In the OED Thesaurus this is  stated by relating an expression to signification, defined as “a thing, event, action, etc., which signifies, symbolizes, or expresses something.” It seems that I have to stop at this very first problem of how “representations” become “expressions” for a while.

Chasing the path of representations

I have always appreciated Daniel Stern’s writing. It is open, honest, and explores its targets systematically. Hence, also the problems in the analysis of mental phenomena stand out clearly. The brief extract in my previous post sent me back to Stern’s “The motherhood constellation” (1995). In the first five chapters of the book a theory of parent- infant interaction is outlined in cognitive terms. It is a masterpiece of showing where the problems of representationism reside. I quote:

Our model insists that maternal representations can influence the observable maternal behavior with the baby; that is, they can be enacted. What would serve as evidence that an effective transformation from a mental model to concrete acts has occurred?

We must be able to describe in concrete behavioral terms how the mother manages to be rejecting and aloof such that her rejection and aloofness can be perceived by the baby and have an impact on him. Can “rejecting” be translated into behaviors such as breaking mutual gaze, putting the baby down when he still wants to be in her arms, or refusing to pick him up when so solicited. Can “aloof” be translated into behaviors such as selectively ignoring some of the infant’s vocal signals or being unable o smile until the baby has done so? (Stern. 1995. The motherhood constellation, pp. 41-42)

…there is a puzzle close at hand: how to conceive of continuity of theme or meaning in the passage from a mother’s representation to her overt interactive  behavior, and then from the infant’s experience of that behavior to his construction of his own representational world. The issue becomes, what are the rules of transformation across these domains.  But the question remains. Can we chase a phenomenon into and out of several domains or levels of description and across the borders of different minds and at the end of the chase be sure we are still following the same phenomenon?  Let us begin such a chase anyway. (Stern, 1995 pp. 42-43)

Stern indeed performed the chase and demonstrated how complex the phenomenon appears to become when it reaches the final transformation, that is, the procedures by which the infant makes sense of the parent’s interactive behavior. I will spare my comments to the next post.