As a point of entry to Professor Simon's admirably learned and sensible article, I should begin with a topic I have at least thought a little about: visual perception and visual cognition. Simon's article suggests that visual perception is like looking at a picture with the mind's eye: the meaning of a scene (or of a picture or a memory image, in his presentation) derives from the associations of that picture. There are several problems with that way of putting things. The idea of the mind's eye seems to me superfluous, regressive, and ultimately nonsensical: if we have a mind's eye, why can't we see our brains? More to the point, it effects a reductionist displacement: visual perceptions, real pictures, and visual memories all become pictures inside the head.
The discussion of mental pictures, though only an aside in Professor Simon's essay, points to a characteristic way that the question of meaning is presented there. That is, meaning, whether visual or (more centrally) linguistic, occurs inside the head, or inside the brain. Visual or language stimuli get plugged in or accessed by the brain, and they evoke images, or associations, or memories there. Meaning, as I understand this presentation, is a subset or sum of such associations, variable with the experience of the mind that is operating.
That's a nicely concise account of meaning, which at least has the virtues of clarity and simplicity. Not very surprisingly, it sounds like a useful guide for the construction of a meaning-simulating machine or computer. A set of relational data inputted into memory is prompted with a cue, and a kind of database meaning can be retrieved in the form of a subset of associated items.
This seems to me a very passive, associationist, and inadequate model of human mentality. I would point to characteristics of another model, starting with an account of visual perception, and based on my understanding of the theories of J.J. Gibson. First, visual perception is not compartmentalized, but is bound up with a totality of perceptual systems in an embodied, mobile subject. Secondly, visual perception is not situated anywhere in particular: not only not in the eye, or in the brain, but not even in the body of the perceiver. That is, perception (visual for the moment) is just the dynamic feedback process of a continuum of afferent information and efferent checking actions. But visual perception is the process, and one could as validly say it takes place in the visual field, or on the surface of the "object" perceived, as in the eye or mind-much less in "the mind's eye."
If this account could validly be applied to other kinds of mental activities, it would suggest a possibly useful alternative way of thinking about meaning: rather than something passively retrieved, or "evoked," as the result of some input or stimulus, we might think of meaning as something dynamically produced, in a continuum of interactions between, say, reader and text, but not located anywhere in particular. That is, meaning doesn't reside in the text, or in the author's mind, or in the reader's mind, but is continuously produced in the process of interaction between reader and text.
This perceptual model has some usefulness and relevance, I think, for understanding how meaning is construed. But visual perception of the environment is quite different from viewing artificial, cultural products such as pictures, and still more different from understanding spoken language, or reading language texts, and we should be clear and careful about distinguishing the processes involved.
Let us skip over picture problems, and look at language issues, since they are usually central to meaning questions, and clearly central to literary criticism. We should further divide language into verbal encounter situations (voiced language acts) and texts. In a situation of verbal encounter, we have something akin to (though by no means equivalent to) perceptual process: a dialogic give-and-take, with situational and nonverbal information added to the meaning cues-expressions, intonations, gestures, body language, gaze avoidance. Situational verbal communication can certainly be deceitful, or inaccurate, or misunderstood, deliberately or not-think of various categories of charlatans, sophists, and rhetoricians (used-car salesman, unfaithful spouses, attorneys-at-law, professors). I would only argue that there is relatively more information about what kind of meaning function is occurring in verbal encounter situations than in a text, and more opportunities for testing-questioning, or expression cues for (mis)understanding or (dis)agreement. Most importantly, meaning in verbal encounter is what emerges in the dynamic give-and-take process: it doesn't belong to one or another party (even if one does all the talking), or to the words, but to the encounter, as what is mutually communicated and (mis)understood.
Language texts are distanced, and offer less specifying and unambiguous information than verbal encounters, or pictures, or perceivable environments (in order of decreasing ambiguity). We might say that meanings of literary texts are so problematic because texts are fundamentally ambiguous; and fictional and poetic texts radically more so than prose, functional texts. Text cues are limited to things like punctuation, or to generic conventions: we expect an instruction manual to make a specific kind of sense, and a literary essay to convey some kind of coherent argument, but a poem or novel can be deliberately non-sensical, fantastic, or absurd, and in any case is licensed to invent characters, and situations, and entire worlds.
What are some implications of these models of perception and cognition for construing meanings in literary texts? I would suggest the following possibilities:
1. Meanings of texts are produced, not evoked or passively retrieved, as part of a continuous process of dynamic engagement with the text;
2. Construing language meanings might be thought of as more like a process of attunement or mutual calibration than like an information retrieval system;
3. Reading, and meaning, are not exactly located: they occur in the text as much as in the mind. We might say that in reading, the mind is engaged in the process of the text. Or even: as much as the text is in the mind, the mind is in the text.