Structural phenomenology: a top-down analytic methodology
Gurwitsch, following Husserl, described two structural parameters applicable to all phenomena: the intensity of our experiences, and their salience, i.e., their experienced relevance to other entities in consciousness. These dimensions subsume experiences within structures indicating the degree of attention consciously paid to phenomena, and their significance to other phenomena experienced simultaneously. For example, the recession to or from unconsciousness of mental contents may be described by the variation of their saliences and intensities. The focal organization implied by these dimensions gives rise to the "searchlight" configuration underlying many models. Consciousness can be structurally analyzed more deeply than this, however. Through incorporation of two other parameters: an internalization of intentionality which I term “microdirectionality,” and a description of the recursive microstructure of the phenomenal field (“layered recursion”), strata of interrelated structures may be employed to explicate experiences in great depth. I will introduce these structural parameters and describe how this more inclusive perspective enables some aspects of both static interrelationships and the dynamics of the creation and dissolution of a variety of sensory, conceptual and linguistic phenomena to be explicated. I will utilize the tip-of-tongue phenomenon as an illustrative example.
Of what use is it to cast description, of any type, in different forms? Surely the answer to the question posed at this level of abstraction is clear; one can merely point to the functionality of mathematical formulations versus, say, verbal formulations of the same facts or theories to see the advantage of particular forms of description in certain realms. What form, then, should descriptions of subjectivity take? I will argue that one may begin to see the answer to this question by considering the work of Aron Gurwitsch (Gurwitsch, 1964, 1966, 1985), who explicates some of the universal dimensions underlying subjectivity. This discovery, I will argue, presents us with the opportunity to cast phenomenological descriptions in terms of structural universals of consciousness. The advantages of such descriptions, then, lie not merely in their potential form, but also in their content: the employment of universals of experiences as the organizing bases of structural descriptions of those experiences.
I claim, then, that a formalism which is thoroughly phenomenological, in structure as well as content, is necessary in order to provide phenomenological description with a symbolism consistent with phenomenal experience. There is a sense, of course, in which any idea is an experienced phenomenon, and so any basis for symbolism is something we experience. However, any structural description in any subject matter is embedded in or superposed over dimensions: characteristics shared by all the descriptors. Thus, in mathematics, to establish a structural symbolism, the number line provides the underlying dimension of quantity, which all functions, and/or graphs plotted or established on it share. Similarly, a structural phenomenological description is one employing universal and independently varying experienced characteristics - thus accurately termed dimensions - to organize subjective phenomena. Of the many models of consciousness or of experienced phenomena, very few which are placed on the firm basis of such universally experienced characteristics. Most such models employ a symbolism based ultimately on an embedding of phenomena in some abstraction of the concept of quantity, usually distance, a characteristic which is not universal to all phenomena, and which in addition implies a precision and/or variables that do not exist.
Thus, in “Morphological eidetics for a phenomenology of perception”, for example, Petitot (Petitot, Varela, Pachoud, & Roy, 1999) presents us with the latest in a long line of elaborate and beautifully worked out mathematical models of conscious phenomena. We see how what is termed “fibration” enables one to symbolize the multiplicities of connotations, or the depth of meaning, contained or implied by relatively simple perceptions. Yet this model, while symbolizing experiences, is not based on experienced universals. That is, the “fibrations” are abstractions of the concept of a mathematical function, created to symbolize multiple functional values. The direct origination of this symbolism from abstract mathematical symbolisms distances it from the phenomenological realm. Phenomena are overlaid on or within this symbolism, but its structural basis is not itself something that we experience. That is, we do not experience, as we perceive or reflect, “a differentiable manifold E endowed with a canonical projection… over another manifold M”, however suitable that symbolism may be in some circumstances to represent mental abstractions. As the basis of a phenomenological description, even as the abstraction of such descriptions, it is, I claim, simply inadmissible.
The same kind of objection holds for constructs such as Fauconnier’s (Fauconnier & Sweetser, 1996) “mental space” model. When we inquire as to the dimensions in which the concepts described are embedded, and how those dimensions interact with the contents, we find that they are merely spatial and temporal separators, derived from ideas of quantity and distance. These are indeed concepts that are capable of being experienced, but they are not universal phenomenal characteristics. Again, the kind of symbolism which must be employed to “contain” experiences must be derived, I am claiming, from those experiences themselves, and must be present in all phenomena.
Of course, most of these models are not intended to be phenomenological models, and so the disparity between experience and the model’s parameters is not surprising. Thus, many connectionist models (e.g., Marcus, 2001, pp. 85-117) of concept structure consist of networks of nodes, each of which, roughly speaking, stands for a concept. Despite the mentalistic orientation of such models, virtually none of their originators would claim that they described concepts as actually experienced.
Gurwitsch’s (Gurwitsch, 1964) model, however, does utilize such a purely phenomenological structure. In that model, patterns of experiences are organized according to universal characteristics abstracted from and present within experiences. This model employs two structural parameters, which according to Husserl, James and Gurwitsch, are characteristic of all conscious phenomena. Those are, first, the intensity with which something is experienced and second, its salience, i.e., its experienced relevance to other elements or entities in consciousness. All experiences possess these dimensions, and they are not mutually derivable, although they may be related, i.e., the most intense aspects of experience tend to be most salient to each other. One finds that these parameters subsume experiences within structures indicating both the degree of attention consciously paid to phenomena, and their significance to other phenomena experienced simultaneously. Such structures, I will argue (following Gurwitsch), are universal, and on their most abstract levels, similar, despite variations in particular contents and persons.
The focal organization implied by this scheme (i.e., a "searchlight" structure) enables us to classify experiences from those most intense/salient to those of which we are barely conscious. Given this model, which might be termed the "basic structural model", one can place the variety of phenomena in consciousness in a field of two dimensions: salience and intensity, and in addition the perhaps more universal dimension of time. This simple searchlight model is the first stage of the structural analysis of consciousness, ably carried out by Gurwitsch (and more recently, by Arvidson). Thus, I (Brown, 2000b, Brown, 2000a, Brown, 1999), and Arvidson (Arvidson, 2000) have argued that these dimensions allow us to organize the contents of consciousness in order to produce general structural descriptions and to make rough predictions. For example, the recession of mental contents to and from unconsciousness may be described by the variation of their saliencies and intensities. We may claim that contents which become less salient must recede, since they are becoming "disconnected," in effect, from the rest of the thematic field. That is, as we attend less to and/or decrease the salience of an experience, it becomes apprehended as, in effect, a less important component of the field of consciousness. Similarly for those contents whose intensity decreases; a lack of intensity implies recession from the focus. The structures elaborated by Gurwitsch, Arvidson, and myself concerning various types of such recessions and transitions result from manipulations of this framework and consideration of its effect on various contents over time, some of which have been explicated by Mangan (Mangan, 1991, 1993) and Galin (Galin, 1994).
One might ask why consciousness is structured in this manner. However, phenomenologically speaking, if all components of the field of consciousness were experienced as equally salient we would not be able to order them, to think and behave appropriately in situations calling for one or a few responses, especially when multiple responses and/or mental contents were mutually opposed. Other, cognitively-oriented explanations have been offered (e.g., see Baars, 1993), but they are not the concern of this essay. The same kind of objections hold for the dimension of intensity. It is necessary that we order thoughts and perceptions on these dimensions, as well as temporally, in order to plan, to respond, and to create conceptual structures. However, despite Arvidson's and others' characterization of various contents of consciousness, and categorization of transitions dependent on the qualitative relationship of those contents and their positions in the above two-dimensional field, we may extend the structural analysis of consciousness so that those categorizations, and others, are more fully explained in structural, rather than content-oriented, terms.
I will argue that the above model can be significantly deepened and generalized by considering two further structural dimensions of experiences. Through incorporation of these two other parameters: an internalization of intentionality which I term “microdirectionality” and a description of the microstructure, within gestalts, of the overall structure of the field (“layered recursion”), strata of directed, recursive structures may be used to describe the constitution of phenomena.
The third parameter, then, is that of recursion. I will argue that the structure of consciousness recurs in multiple layers within its components, i.e., that by considering the repetition within gestalts of the overall structure of the field, layers of directed, recursive structures may be used to describe phenomena. With the addition of these further elaborations, a structural description of the field of consciousness may present at least partial descriptions of and some explanations for, among other things, the formation and dissolution of cognitive objects, the greater or lesser articulation and elaboration of ideas, and some linguistic phenomena. Finally, the fourth parameter, which I will term, generally, "directionality” (and when related to low-layer structures, "microdirectionality") characterizes the experienced orientation, both temporal and atemporal, of one gestalt toward another. This is, roughly, a consideration of what experience we are "led to" by the one we presently focus on.
My justification for the parameter of recursion is as follows. Gurwitsch, James, and others characterize the field of consciousness as unified, i.e., gestalt in nature. This field has a particular structure, described above. However, the individual contents of this field, although interconnected within it, are themselves gestalts, i.e., fields unifying separate components. In order for those contents to be experienced coherently, it is just as necessary that those contents be organized as it is for the field as a whole (see below). Since the contents are unifications of components, those components have the same general characteristics as do the gestalts of which they are elements. Thus, any particular gestalt of which we are aware must itself share the "searchlight" structure of the overall field of consciousness. This is a perfectly general argument to the effect that just as the whole experienced field must be structured through the dimensions of salience and intensity in order to separate, organize, and relate its contents, so must any individual gestalt, for the same reasons. Reflection does indeed confirm that our experience of the table in front of us, for example, possesses an area of greatest intensity and salience; yet as we are, in addition, aware of the table's surroundings, we must realize that in order for those latter contents to be organized and experienced as objects they must be similarly structured. That is, the floor under the table, although we are in effect only peripherally aware of it, must also have a focal area organized, in general, just as the table is organized. This focal area might consist of the concepts of flatness and hardness, rather than, for example, the thickness of the floor.
Therefore, aspects of the table must in their turn share such organization, or be in danger of having, in effect, no structure, i.e., having amorphous internal organization. Yet this latter cannot be true, for two reasons. First, we are aware that they are particular objects, and second, that particularity implies such organization; that is, if all the components of, say, the floor, were simultaneously and uniformly presented, we would not be able to tell whether it was a floor, a flat colored surface, an object with no mass, or a skin of some indefinite composition. All these are possibilities and contents subsumed within a floor, but restrained, in effect, through their intensity and salience relationships. Thus those must be preserved even when the object is not thematic. The same considerations hold for the field of consciousness itself. Bernet, in fact, mentions that Husserl speaks of “a gradual constitutional analysis in which the unities already constituted within a continuous multiplicity function in turn as multiplicities constituting a higher unity” (Bernet, Kern, & Marbach, 1999, p. 139). Gurwitsch, in his brief analysis of what he terms the “outer horizon”, touches on phenomena in which, for example, the “unseen side” of a building “is referred to only in its contribution towards the architectural configuration of the building” (Gurwitsch, 1964, p. 373).
If the above is true, then the components of gestalts, gestalts themselves, possessing an internal structure organized in accordance with the principles organizing the field as a whole, are aspects of one huge recursive structure. This is a crucial realization: the structure of consciousness is recursive. There are several immediate implications of this structural recursion. First, we may now look to various depths of this structure for the explanations of a variety of phenomena. That is, I will term the components of the primary theme – the experience which we are most focused on - "connotations”, while components of those connotations will be termed "microconnotations” without regard to the level on which they are experienced. Second, we have now an approach to the characterization of the field as atemporally continuous (Gurwitsch and James, for example, considered only temporal continuity). Consider, for example, the number line. This is revealed as continuous because (among other reasons) of Dedekind's definition of neighbors: no matter how close to some point another is, there is always a point between the two. Similarly, given the recursive structure of consciousness, we might speculate on an analogous characteristic of this field, but accomplished within "layers" of contents. Thus, it is possible to say that no matter what layer we are considering (i.e., what layer in the recursive structure we are experiencing or describing, either in the moment or reflectively) in the field of consciousness, there are always "microconnotations" which are components of the next layer of recursion below the present one, which comprise the connections between connotations on the layer above – the continuity. The recursive structure has provided an analog to the continuity of the number line. This supplies a structural explanation for the continuity of the field of consciousness. The limited ability of one's capacity to "hold things in mind" simultaneously can be conceptualized as a limitation of the components of any given layer, while allowing for virtually infinite connotational content in depth. On the topmost (and perhaps any other) layer of this structure, there are a finite (and surely a decreasing number as we ascend) number of gestalts that the field of consciousness may contain. Yet as we progress downward through the recursive structure, the number of gestalts, while (perhaps) still finite, increases with each layer, although they may also become more faded and blended (see below for the analysis of those terms) as the recursive depth increases.
Let us consider further the nature of transitions. We may now characterize transitions between gestalts, not merely through the relatively gross conception of “relations” - even when those relations are classified into types - but in finer analysis, through "microconnotational bridges”. Thus, webs of connotations produce the richness of meaning not only within gestalts, but connect two (or more) gestalts through layers of (directed – see below) microconnotations. In other words, a relationship is itself a gestalt, with its own connotations and structure mirroring, again, the overall structure of the field. Thus it must have microconnotational components, and they must aid in the act of creating the relationship between the gestalts that particular gestalt bridges. Thus, in particular, if the floor is under the table, we might erroneously characterize the relationship of being under as unitary, i.e., opaque and unanalyzable. But this cannot be true, for in this example, we are experiencing a table and a floor, not, for example, a piece of toast and a plate, or even the table and floor in a dollhouse. Those latter are also characterized, in part, by the relationship of being under, yet that relationship must be different for those gestalts. To take only one example differentiator, the distance which one thing is under another varies in each situation. How is this difference realized, and how may it be explicated? It must be that the relationship of being under must have its own connotations; yet as we have seen, this implies that those connotations must also be in the same structured relationships as the field of consciousness. Yet in addition that relationship, being under, must also, since it is situationally specified (i.e., relates specifically to the particular objects and their surround), contain within it as connotations part at least of the theme of which it partakes, i.e., those objects: the table and floor. But the gestalt, being under, then, containing the table and floor which, on another layer, it relates, must at that micro layer interrelate, i.e., constitute the relationship between, those two contents. If this is true, then that relationship, i.e. the web of microconnotations within the being under gestalt, since it governs the being under gestalt, must be experienced as an aspect of its highest layer of structure. There is then a very real sense in which the relationship bridges the gestalts which it relates.
This bridge then must be at least in part composed literally of interconnected webs of microconnotations (which because of their own microconnotations, are a continuum). Now, considering the nature of this linkage brings us to the next structural dimension, that of “directionality”.
When one considers the internal organization of gestalts, one realizes, then, that not only is the gestalt structure recursive, but it is internally directed. That is to say, if we have as the theme a tabletop, we will tend to be aware of the table legs as necessary implications (and thus connotations) of that tabletop; the legs are, in a sense, contained (recursively, as we have seen above) within that tabletop. The vase on the table may be equally salient, even more salient, if new, for example, than the table legs, yet there is no necessary implication of and presence within that table of the vase, as there is of the legs, i.e., no directedness from table to vase as there is from table to legs. The phenomenal tabletop refers us toward its legs, for example, preferentially to objects external to the table. I will term this "microintentionality," or alternatively, "microdirectionality." Notice that this is not necessarily (although I will argue below that directionality includes) a temporal transition. Thus, we can experience two types of microintentionality, an atemporal, as in the above example, and a temporal, in which connotations shift in accordance with implications and associations.
Husserl, in fact, seems to have been aware of this parameter of consciousness to a certain extent. Thus, he wrote of “interconnections among intentions, that is, indications which at any given moment run… from phase to phase; one is drawn forward within the familiar complex in a line of constant belonging together” (Bernet et al., 1999, p. 128). Similarly, he speaks of “concealed intentional implications” in judgments (Bernet et al., 1999, p. 201). It may also be that the mention of “configurations of sense” which “point back within themselves” (Bernet et al., 1999, p. 201) relate to this dimension. In addition, Gurwitsch mentions that various “facts” appearing in what he terms the “margin” of the field of consciousness “are not experienced as scattered data but.. as pointing and referring to each other, and thus pertaining to a coherent order” (Gurwitsch, 1985, p. 78).
One might question whether directionality is truly an independent dimension. Is this not merely a characteristic of salience? That is, is salience intrinsically directional? Yet one can come up with examples in which this does not seem to be the case. Let us take, for example, the table on the floor. Suppose that these were part of the contents of a house, and the theme of the field, the focus of our attention, was on the rug in that particular room. Now in that case, the salience of that rug, on the floor, under the table, is the highest of the objects in that house. Yet it must be, nonetheless, that the interrelationship between the table and its legs, or its top, must be of a particular sort which privileges those components in the context of those gestalts. As the table is seen, or understood, even peripherally, its legs and top are necessarily implied through that experience, even though we are interested primarily in the rug on the floor, even though the legs of the table are on that rug and we are oriented more toward those legs than to the tabletop. The legs are part of the table, and we know this as we see (or contemplate) them. The table as object cannot dissociate even though we focus on the rug beneath its legs. Thus, we must consider that there is indeed another dimension in addition to that of salience: directionality or microintentionality.
Let us take another example, that of the seats versus the backs of objects on which to sit. Given stools, etc., we see that while a seat is essential, a back is not. Thus there is an asymmetry in the directionality of this abstract object: the seat may refer to the back, but the back must refer to the seat. This example indicates, again, the independence of salience and directionality; since both seat and back may be equally or differentially salient, but the above asymmetry holds.
The implications of microintentionality are far-reaching. If it is true that connotations and microconnotations are organized not merely in salience relationships, but that those salience relationships are directional, i.e., that given two microconnotations of equal salience one might well predominate as most necessary (in some situation) and thus tend to attract temporal transitions and in addition color the gestalt in a particular way determined by its own microconnotations rather than the other's, then we must conceive of the field of consciousness as thoroughly directional throughout. It is possible to conceive of explicating the Husserlian phenomenon of "running-off" (Husserl, 1990) through an analysis cast in terms of shifts of intensity and salience in lower layers of microconnotations so that the same gestalt can be presented as more or less “faded.”
In addition, we may conceive of objects of all sorts in the field of consciousness as arising, forming, and dissolving through processes of formation and redirection of microconnotations interrelated through microintentional "directionalities." The table above is an object precisely because its components are inter-referring, and will maintain that status as long as they are. That is, one may conceive of objects (and particularly non-sensory objects, since sensory objects are maintained through very low-level processes in the CNS) as being generated through inter-referring directionality, so that the set of connotations (given recursion, the field) within an object inter-refers preferentially to the other connotations of that object, while referring less strongly to connotations of other objects.
In this context, I have introduced what is, in effect, a vector field; that is, the field of consciousness, given all the above, is something like a recursive vector field. Yet we must ask just what the term "vector" can mean here. We cannot speak of numerical magnitude, as experienced, nor can we speak of any absolute coordinate system within which to embed directions. Indeed, vectors are, at base, a conception designed to describe an aspect of reality in which a correspondence to numbers can be assumed, and vector operations are performed on numbers, all of which have the same qualitative nature. In other words, one can operate on numbers with proscribed operations; one cannot operate between entities whose qualitative natures are different without a previous mapping of the interrelationships of those entities. However, the analogy to vectors holds remarkably well: if we consider the nature of objects, above, the internal inter-referral intrinsic to their structure lends itself to being modeled by a vector field of the type employed to describe magnetic fields in physics: the del X field. Yet there can be, it seems, no meaning to the corresponding vector operations, since the qualitative differences between the microconnotations preclude numerical correspondence and thus numerical operations.
The four parameters above can now be applied to analyze some specific experiences. I will start with tip-of-tongue (TOT) phenomena. We have seen (Brown, 2000) that, utilizing only the parameters of salience and intensity, one can describe the recession of experiences to the fringes of the field of consciousness. But while decrease in salience and/or intensity describes this recession in simple terms, a deeper description is possible if we consider the microdirectional and recursive aspects of the field.
Suppose we consider some content of consciousness: the gestalt comprising our current experience of some person, say: their name, face, personal particulars, and so forth. This gestalt is a unified whole, with inter-referring microdirectional vectors, layers of microconnotations, and so forth, as described above. If we forget that person’s name, we may describe that situation in one of two ways; through what might be termed “recursive recession” or through what might be termed “microdirectional reorienting”. Recursive recession occurs when a connotation moves to a lower layer within the recursively structured field within a particular gestalt, as when, for example, the hardness we experience as a component of a tabletop becomes an aspect of the paint on that tabletop rather than of the tabletop itself. Thus, if that forgetting has occurred by means of recession down layers of recursion, into what we might term the “microconnotational background” of that particular gestalt, then the directional references within that gestalt will still point toward that component. In other words, microdirectional vectors still point towards it. However, since it will have receded as it sinks through layers of connotations, it will be much more difficult to focus on because of it’s own lack of intensity and because of interference from other connotations at higher levels. Thus, we experience a “gap” in that gestalt; something like a “hole” or “sink” whose bottom we move toward but cannot find. If, on the other hand, the directionalities, instead of being "pointed" toward the gap, were pointed away from it, toward the connotations already present, then the gestalt would have repaired itself, sealed up the hole, so to speak, and there would be no TOT experience. I will have more to say about this below.
The second type of forgetting might occur if that same component, instead of receding within the gestalt, instead becomes less referred to, i.e., less integrated into the gestalt. That is, if there is a reduction of “micro-salience”, i.e., salience within that gestalt, resulting from a decrease in the number of microintentional vectors referring to that component, it separates from the gestalt. If the microdirectional field (the microintentional inter-referring) ceases to include that memory, it becomes detached from its original context, and if that original gestalt remains in the foreground, that memory recedes to the fringe of the field of consciousness. In this case, the gestalt has no “hole”; it simply does not refer to that component any more, and we forget that it had been there; the gestalt re-forms without it. In this instance, there is no TOT phenomenon, and no experience of a gap.
In summary, either the missing component might recede to the "edge" of the whole field of consciousness, "outward"; or it might have receded "down", so to speak, i.e., to lower layers of the recursive structure of that particular gestalt. One might expect the former to be the case when the missing memory was a gestalt "in parallel," in effect, with the original gestalt, as with an item of clothing. An item of clothing associated with a person is nonetheless a complete unity in itself, and might be associated with other people and contexts. Here we might expect the phenomenon of microdirectional reorienting to occur, and no TOT experience. If the connotation of the original gestalt was one in which the memory was of some intrinsic component of the person: their face or name, we might expect a TOT experience to be more likely: a recursive recession, rather than a microdirectional reorienting. Given these two classes of possibilities for the missing memory, the microdirectionalities may either be pointing inward or outwards from the original gestalt.
Is it possible to have a TOT experience with microdirectional reorienting? If enough directional vectors pointed to what is now another gestalt, and that latter had receded far enough into the fringe, that would seem possible. In that case, we are looking for a memory which is not strongly connected to the original gestalt, but which is experienced as having some connection. But we would not experience this as an attempt to recall an aspect of the original gestalt but as the attempt at recalling something vaguely connected to it.
One may also generalize the TOT experience to non-noun and even non-verbal TOTs: “What was I going to say?” in the case of an idea. This of course starts with part of the idea, or with a general formulation of the idea, and then requires the specifics, not necessarily even verbally. Second, the recall of arguments, i.e., reasoning to some conclusion, “I can’t remember how I got this, but I know I did…” Again, the general idea of the flow of the argument, or some key idea in it evokes the whole. Third, “What was I doing here, looking for, or about to do…” The recall of some previously planned action sequence given part of it or a general idea of what it was. Thus the memory missing in the TOT situation may be an aspect of a static gestalt, as in the case of a face, a dynamic complex, as in the case of a forgotten argument, or an idea in parallel to the original.
Now let us address the dynamic aspect of gestalts. Given the above, the implication is that a concept is formed when a complex of gestalts is preferentially inter-referring: when the microdirectionalities of the complex refer toward the other elements of the complex rather than to connotations or microconnotations outside of the complex. This complex, then, becomes differentiated from other contents of consciousness by virtue of that structure. We can see that concepts may form, dissolve, combine, or dissociate depending on the microintentional flows. Now what is it precisely that directs those flows? Surely we cannot say that something called “similarity”, even “functional similarity” causes them, or even describes them. What is similarity, after all, except some merely arbitrary and transient characteristic or function?
We might, however, get some idea of the processes underlying gestalt dynamics if we consider a phenomenological explication of the differences between concrete images and symbolic abstractions. That is, we can start by observing that the idea of a cat, for example, is differentiated from the background by the same processes of microdirectionality that differentiate more abstract concepts. Why then is it so much more definite and stable than some more abstract idea? Because this former idea interacts with actual cats, pictures of cats, etc., in our experience. Thus every time we see a picture of a cat or interact with an actual cat, that concept is reinforced: structured and enriched, and consequently becomes more differentiated (containing more internally inter-referring microintentional vectors) and recursively deep (containing more – or more accessible and varied – layers of microconnotations). That is, that complex, which is simultaneously being further isolated because of the discrimination and aided in its enrichment by further perceptual interaction, is consequently “gathering” to itself, so to speak, other gestalts: further implications, connotations, etc. Thus the complex grows both in “width” and in “depth”, and this growth and isolation, i.e., its differentiation from other concepts, is aided as we further interact with cats.
A more abstract concept, on the other hand, cannot be so easily maintained through direct perceptual interactions. However, our related interactions: verbal, ideational, social, serve to reinforce and enrich that concept in a manner generally similar to the more concrete concepts: causing the further differentiation of its microintentional flows, just as the more concrete one does. However, since abstract concepts are less easily and directly reinforced in this way, they are less stable and definite, or at least organized around a less definite core than an image (whether visual or sonic, etc.). Thus one would expect that abstract concepts are more easily formed, reformed, dissolved, altered, etc. The concept of freedom, for example, is highly variable and flexible for just these reasons, and may even disappear, or virtually so, into the background.
A concrete image, then, may in many circumstances be regarded as a “source” of microconnotations in something like the vector sense, i.e., a vertex from which microdirectional vectors point from and from which microconnotations originate. The development of the possibilities of the concrete in terms of the enrichment described above would seem to imply that the directionalities point outwards from this source, since a microconnotation generated from a connotation already present would point toward that connotation. The reverse might not be true to the same extent, since the connotations would primarily focus on a subset of the microconnotations, given the recursive “searchlight” structure of the fields.
The inverse of this, a “sink”, might occur when connotations were disappearing, as when we dismiss erroneous retrievals in a TOT situation. That is, the process of actively inhibiting (e.g., Anderson & Green, 2001), “discarding”, an idea causes that idea, with its associated connotations, to disappear, in some cases, at least, through recession (Brown, 2000b) to the fringe of consciousness. That idea, then, becomes a kind of “hole” in the field of consciousness, and should drag, so to speak, its connotations with it.
These possibilities are, then, a few of the logical consequences of the structural implications of the more complete model of the field of consciousness.
We might also briefly speculate about the phenomenological structure of language. Suppose we needed to manipulate concepts representing concrete perceptual objects in extremely flexible ways. If we wanted to think about cats, say, and had to employ concepts which were concrete: a visual image of a cat with associated connotations, for example, we would, as a consequence of its being centered on an image, be severely restricted in how we could manipulate it and interact that concrete concept of cats with other concepts. Manipulations that were not compatible with the special characteristics of that visual image would be extremely difficult and disconcerting to carry out, it would seem, since the perceptual basis of the image and its implications, e.g., the kinesthetic, impose enormous stability on such a concept, and could be incompatible with various thoughts about cats. Suppose, however, that we, so to speak, “hollowed out” the focus of that gestalt, and replaced that image, or whatever constant stabilizing gestalt lies at its focus (and relegated that image to the periphery rather than the focus of the gestalt), with something neutral, whose manipulation or utilization in different contexts had no strong relationship to that entity. An arbitrary symbol, like a word, would be ideal. The connotations of the image of the cat would still be present, in the main, yet the exact visual image and its connotations would be missing (or much less salient), allowing much more flexible manipulation of that concept. This, if true, might be the beginning of a phenomenological approach to language.
Given the above, the dynamics of microintentional structure would seem to have almost the opposite relationship to the phenomenon of metaphor. That is, suppose that instead of a concrete image (visual, aural, etc.) being the cause of connotative limitations or rigidity, we consider it as the possible origin of a desirable stability. More specifically, if the argument above is correct, concrete images must be more stable than abstract concepts. A concrete image, then, might be employed as a source of stability, as well as a source of particular contents, for an abstract concept. In other words, metaphor might be the source of a particular set of connotations which one wishes to conserve. If one forms or originates an abstract concept and desires (on some level) both particularity and stability for that concept, then a possible mechanism could be to “borrow” microconnotations from a related concrete image, or if there are no available or obviously related concrete images, to take one reasonably related and use that as a source of stability. We might speculate that this is in part the phenomenological function of metaphor. Taking a specific example, let us consider at the abstract idea of love. Purely as an abstraction, this idea is extremely vague, and may evoke no more than a few feelings, and perhaps some images from a love relationship. However, if we apply the metaphor, “love is a journey” (see Lakoff, 1990 for an elaboration of this metaphor), borrowing concrete images of landscapes, movement, vehicles, and so forth from memories of journeys, the abstract concept of love, partaking of these, becomes richer, more concrete, and more stable. At the same time, the price we pay for this stability and richness is a certain rigidity: love now is a journey for us, and it contrasts to someone else’s conception of love as, say, a contract.
What has occurred, dynamically? Rather than speaking of “mental spaces” (e.g., Fauconnier & Turner, 1996), we may conceive of a kind of superposition of vector (directional) fields, where, when the two gestalts: love and journey, are juxtaposed, multiple and complex operations take place. Is a third concept, a joining of subsets and implications of the two originating ones, formed? Are the two joined so that one, at least, “borrows” from the other? It seems to me that either of these, and several other alternatives, might take place, depending on the context in which the metaphor, the juxtaposition of these concepts, occurs, and on the contents of the initial concepts. Indeed, even if, initially, a third concept (love-as-journey) is formed, this might be unstable and “collapse” into the concept of love, to ultimately replace it. Given the structural picture in this essay, we are able to conceive of these various alternatives without basic theoretical conflict.
This approach, then, is the beginning of an explication of concept formation and language in general phenomenological terms, made possible by the structuring of consciousness by the four parameters described above. It is a kind of mentalistic “meaning is use”, in the sense that the use of a concept, as it relates to other concepts and to the world, refines and selects, in a somewhat Darwinian sense, the components of that concept. For example, we might envision one of the underlying drives as the necessity of providing stability with adaptability. In this sense, then, on the basis of such criteria as their internal and external functioning and compatibility, concepts are intrinsically dynamic, changing with (or without) use in goal-oriented ways (yet to be described).
In the above description, a non-formally recursive method of generating a structured continuum, there is a sense in which the individual components are unimportant and what is important is the result of the “vectorial” combinations of their microintentional flows. This combination, by whatever processes realize it (and these may vary), results in a field in which areas without definite boundaries, corresponding, perhaps, to attractor basins, establish themselves within a dynamic complex of such microintentional flows. It is these stable areas, then, rather than any of their explicit and demarcated components, which constitute the functionally important aspects of this field, and it is the study and explication of the general principles of their formation and dissolution which at this point constitutes the aim of this model. Hence, employing operations which might correspond to such PDP protocols as "spreading activation", we might think in terms such as altering field density and directions, so that varieties of sources and sinks interact dynamically, in combination with movement through the field layers and into and out of the dynamically altering boundaries of concepts.
Why specify, directly or indirectly, contents at the microconnotational level? One might imagine this approach to be equivalent to a set of explicit lists that provide vector elements specifying the elements at the higher levels. However, this is not my intent. There are several differences between specifying content in terms of lists of features, however hierarchical those lists are, and the present approach. First, in a list of features, the features are what Fodor (Fodor & Pylyshyn, 1988) has termed “punctuate”, i.e., separate, atomic, and not related except by their mutual presence in the particular list and perhaps by within-list ranking. In the above approach, in contrast, the “features” are not atomic, since each itself consists of a network of features, and thus unitary elements are considered such only for descriptive convenience. In addition, since they are interrelated through microintentional vectors with their own microstructures, they are formally continuous: they are, considered as a whole, no more features than numbers are features of the number line. Yet each recursive level, considered in isolation, may indeed be described in terms of features, in contrast to the number line. In addition, the notion that the features described here can be replaced by, or are equivalent to, words is in the main erroneous. Connotations may be words, images; in fact, any kind of classically conceptual or non-conceptual mental entity. Thus, connotations are not descriptions in a logical sense, and words, as descriptions of connotations, must be considered symbolic abstractions of the underlying phenomenal experiences.
In addition, I do not believe that compositionality, in the sense that Fodor (Fodor, 2001) uses the term, applies here in quite the way he defined it in his essay. On the one hand, it does apply to the extent that experiences do have components, and that those components do supply contents to concepts. On the other hand, since I consider the components of experienced phenomena, and the phenomena themselves, to be gestalt in nature, the addition or subtraction of components to these field-structures alters both the field and the component, to some extent. Thus compositionality is context-sensitive, not strictly additive, and the meaning and references of “host” concepts may or may not be compatible with the meaning and references of their components.
However, given that the relations between gestalt components are complex, there is certainly room for operations which might be considered “syntactic” or blends of semantic and syntactic, either on conscious or unconscious levels. It must be noted that this model is exclusively of consciousness. It is clear that there are unconscious processes involved in the generation and modification of virtually all aspects of conscious experiences. On that level too, then, there is room for a variety of operations, which may equally be considered mental or purely neural for the purpose of this model. These latter operations, which have only begun to be elucidated, may produce a variety of effects on the field of consciousness.
Anderson, M. C., & Green, C. (2001). Suppressing unwanted memories by executive control. Nature, 410, 366 - 369.
Arvidson, P. S. (2000). Transformations in consciousness: continuity, the self and marginal consciousness. Journal of Consciousness Studies, 7(3), 3-26.
Baars, B. J. (1993). A cognitive theory of consciousness (3rd ed.). Cambridge, MA: Cambridge University Press.
Bernet, R., Kern, I., & Marbach, E. (1999). An introduction to Husserlian phenomenology. Evanston, IL: Northwestern University Press.
Brown, S. R. (1999). Beyond the fringe: James, Gurwitsch, and the conscious horizon. Journal of Mind and Behavior, 20(2), 211-227.
Brown, S. R. (2000a). Reply to Bruce Mangan's commentary on "What Feeling Is the 'Feeling of Knowing?'". Consciousness and Cognition, 9(4), 545-549.
Brown, S. R. (2000b). Tip-of-the-tongue phenomena: an introductory phenomenological analysis. Consciousness and Cognition, 9(4), 516-537.
Fauconnier, G., & Sweetser, E. (1996). Spaces, worlds, and grammar (1st ed. Vol. 2). Chicago, IL: The University of Chicago Press.
Fauconnier, G., & Turner, M. (1996). The many-space model of conceptual projection.
Fodor, J. A. (2001). Language, thought, and compositionality. Mind & Language, 16(1), 1-15.
Fodor, J. A., & Pylyshyn, Z. W. (1988). Connectionism and cognitive architecture: a critical analysis. Cognition, 28, 3-72.
Galin, D. (1994). The structure of awareness; contemporary applications of William James's forgotten concept of "the fringe". Journal of Mind and Behavior, 15(4), 375-402.
Gurwitsch, A. (1964). The field of consciousness. Pittsburgh, PA: Duquesne University Press.
Gurwitsch, A. (1966). Studies in phenomenology and psychology (5th ed.). Evanston, IL: Northwestern University Press.
Gurwitsch, A. (1985). Marginal consciousness. Athens, OH: Ohio University Press.
Husserl, E. (1990). On the phenomenology of the consciousness of internal time (Vol. IV). Dordrecht, The Netherlands: Kluwer Academic Publishers.
Lakoff, G. (1990). Women, fire, and dangerous things (2nd ed.). Chicago, IL: The University of Chicago Press.
Mangan, B. (1991). Meaning and the structure of consciousness: an essay in psychoaesthetics. University of California at Berkeley, Berkeley, CA.
Mangan, B. (1993). Taking phenomenology seriously: the "fringe" and its implications for cognitive research. Consciousness and Cognition, 2, 89-108.
Marcus, G. F. (2001). The algebraic mind: integrating connectionism and cognitive science. Cambridge, MA: The MIT Press.
Petitot, J., Varela, F. J., Pachoud, B., & Roy, J.-M. (1999). Naturalizing phenomenology: issues in contemporary phenomenology and cognitive science. Stanford, CA: Stanford University Press.