I will start with the SGP thesis in Fink’s terminology—the thesis that there can be “solely generic phenomenology”. An example would be the experience of something as red without an experience as of any specific shade of red. Fink says I am forced to accept solely generic phenomenology but that it “has been introduced to argue against Block” (p. 10).
Fink is talking about my “overflow” arguments. These arguments are based partly on an experiment by George Sperling (Sperling 1960) that is covered in all respectable introductory psychology courses and on experiments from Victor Lamme’s lab in Amsterdam that combine Sperling with “change blindness” (Lamme 2003). The upshot of these lines of research is that there is more “capacity” in phenomenology than in cognition—or so I have argued. In the Sperling experiment , subjects are presented briefly with an array of letters, for example, 3 rows of 4 letters. Subjects say they see all or almost all of the letters but when asked to name the letters they saw after the stimulus has disappeared they can name only 3 or 4. Sperling’s innovation was to “cue” subjects to report one specific row after the offset of the stimulus, using a tone to indicate the row. The finding is that subjects can report 3 or 4 items from any given row, supporting the idea that their phenomenal “iconic” representation really did include information about the specific shapes of all or almost all the letters in the array.
In Lamme’s experiments the subject is briefly shown an array of, for example, 8 oriented rectangles. After the array is turned off, the screen goes dark for up to 4 seconds, then there is a second array of 8 rectangles in which one of the rectangles may have a different orientation (e.g., vertical rather than horizontal). A cue—a line pointing to the location of one of the rectangles—can occur when the screen is dark or, alternatively, when the second array appears. The subject’s task is to say whether the rectangle in the cued location has changed orientation. If there is no cue or if the cue comes after the new array has overwritten the iconic representation of the first array, subjects have a capacity of about 4 items. But subjects say they have a kind of image of the array in the dark period after the stimulus has gone off. When the cue is presented in the dark period (up to 4 seconds later in some versions) after the stimulus has gone off they have a capacity of up to 7 of 8 items. I have argued that this pattern of results indicates that subjects have a persisting conscious mental iconic or imagistic representation of 7 of the 8 rectangles of sufficient specificity to compare orientations of the initial rectangles with the final display of rectangles even though they can “cognize” only about 4 of them—in the sense of storing them briefly in “working memory”. The upshot according to me is that there is more capacity in the phenomenal iconic representation than in cognition and thus that phenomenology “overflows” cognition.[1]
The subjects report seeing all or almost all of the items and the cuing experiments—showing as they do “partial report superiority”—appear to back up what the subjects say. However, as Fink notes, my opponents[2] criticize my appeal to what the subjects say about their experience (Byrne et al. 2007). What is in the subjects’ consciousness might be just a generic representation—e.g., indicating that there is a circle of rectangles or array of letters without indicating the specific orientations of the rectangles or specific shapes of the letters. After all, we can’t expect naïve subjects to have a grip on the distinction between generic and specific phenomenology. Subjects say they have an image of all or almost all the items because they have a solely generic representation i.e., a representation that specifies the location of the items and their abstract category (letters, rectangles) but without the specific details (letter identity, orientation).
How do my opponents explain the fact that the subjects can get 7 of 8 rectangles right and 3 to 4 letters from any cued row if their phenomenal icons do not contain the specific information needed to do these tasks? According to these opponents, the specific details of the shapes are registered unconsciously. And when subjects think they are reading details off of an already present conscious image, what they are really doing is making unconscious details conscious (Phillips 2011). Fink concludes that “Allowing SGP thus blocks Block.” (p. 10)
My response to Fink consists of three points: (1) my argument for “overflow” does not require any blanket rejection of solely generic phenomenology. (2) I have not issued any such blanket rejection and I have given qualified endorsement of some kinds of solely generic phenomenology. (3) I think there are some crucial cases—notably some spatial geometry cases in which there is reason to doubt solely generic phenomenology. I will explain these points.[3]
Why does Fink suppose I cannot accept generic without specific phenomenology? Part of his argument is that for an imagistic representation there cannot be generic without specific phenomenology because images are “concrete”. He says
Imagistic representation… does not allow for such indeterminacy, because images exploit the isomorphisms between concreta….Introspectively, our phenomenal experiences resemble images. If phenomenal experience represents image-like, then there cannot be SGP—and p-precision seems dangerously close to being trivial; if phenomenal experience is non-imagistic, then we can allow for SGP and render p-precision non-trivial—but this is in tension with some of Block’s other work and our introspective evidence. (p. 10)
A similar argument to his was made by Robert Van Gulick (2007) and in a different form by Rick Grush. Van Gulick says
If one holds a “movie screen of the mind” model of phenomenal consciousness, it may seem impossible that there could be letters that are phenomenally present as letters without being present as specific letter shapes. But such a model is at best problematic, and if one rejects it, then there seems no reason why the characters of which the subjects are aware could not be indeterminate in ways that exactly match their limited cognitive access to those features. (p. 529)
In my 2007 reply to Van Gulick I rejected this argument and—contrary to what Fink says about my argument—I endorsed a version of the SGP. I said
Van Gulick notes that the “movie screen of the mind” view would say that you cannot have generic phenomenology without specific phenomenology, implicitly suggesting that I am relying on the “movie screen of the mind” view, and on the fact of generic phenomenology, to argue for specific phenomenology…I reject the principle – applied by … Van Gulick – that pictorial representation has to specify the relevant details. I call this principle the “photographic fallacy” (Block 1983). More specifically, the photographic fallacy supposes that pictorial representations have to represent details of anything in view in the manner of a prototypical photograph. To see the fallacy, note that an impressionist painter might represent a hand in broad brush strokes that do not explicitly represent the number of fingers or whether one of them has a ring. (Block 2007b, p. 533)
It may be said that endorsing generic without specific phenomenology on my part is just incoherent since it undermines my own position. Recall that the reason Fink says I cannot endorse generic without specific phenomenology is that my opponents use it to argue that what is in consciousness in the Sperling and Lamme experiments is solely generic, the specific details being perceived unconsciously. My approach has not been to issue a blanket denial of the possibility of solely generic phenomenology but rather to argue against the claim that the highly specific representations in these experiments are unconscious (Block 2007a, 2011, 2014b).[4]
Is solely generic phenomenology possible? There certainly are some intuitively plausible (though not compelling) cases. For example, if one sees a red thing in the distance one may perhaps see it as red without seeing it as having any specific shade of red. (See Stazicker 2011, forthcoming for defenses of solely generic phenomenology.) However, even if there is generic phenomenology, I think it is doubtful in some cases, notably certain spatial cases. In particular, I doubt that there can be generic phenomenology of an oriented rectangle that does not specify the rough orientation of the rectangle.[5]
My rationale for this view is partly introspective and partly a result of informal reports of imagery experiments from Stephen Kosslyn. I have discussed doing experiments on this issue with Kosslyn and Dan Reisberg.
Imagine that you are in a house, going down the stairs and out the front door. In front of you is a picket fence with a gate. You go out through the gate and walk to the corner where you mail a letter.
Stop now and answer the question: which way did you turn when you went out through the gate? Kosslyn reports in conversation that when he gives such spatial vignettes to subjects they do not report that there was no particular direction. The experimental challenge is to design an experiment that distinguishes between an answer made up on the fly and an answer based on what was “already there” in the image.
To summarize so far: Fink says “Allowing SGP thus blocks Block.” I reply that my argument for “overflow” does not require any blanket rejection of solely generic phenomenology; that I have not issued any such blanket rejection; that I have endorsed the possibility of solely generic phenomenology; and that I think there are some specific cases in which solely generic phenomenology is not very plausible.