Referee #3 anon

The present article is quite a difficult one to review. I suspect that referees with expertise in Philosophy have already been contacted, but I would strongly encourage you to contact experts in space/form perception, such as Jan Konderink, James Todd, or Ennio Mingolla. Researchers in these fields could more adequately evaluate the theory advanced by the author. I provide my own review below.

Lehar's manuscript has great potential to become a BBS target article. IN IT'S PRESENT FORM, HOWEVER, I FIND IT INADEQUATE for publication. Below I outline some of the reasons.

It seems to me that the author is attempting to do "too much". On the one hand the author seems to engage in a philosophical battle with several century-old issues. At the same time, the author proposes a novel, original theory of space perception. It seems to me that the greatest contribution is by far the latter. The philosophical arguments should take part of background material, or discussion, but in a much more summarized manner. They serve to motivate his theory. I find the present "AGGRESSIVE TONE" INADEQUATE, AND TO BE FRANK, NOT VERY PERSUASIVE.

[Note: Caps above indicate sections highlighted by the editor for particular importance to address.]

Much of the philosophical discussion to which this reviewer objects was originally not part of the paper, and was added by demand of Reviewer #1 in the first round of review. I agree with this criticism, consequently I have removed some of that added material.

As for the "aggressive tone" which this reviewer finds "unpersuasive", this is an attitude I have developed as a consequence of a series of encounters with reviewers like this one. He says he is "not persuaded" but does not bother to explain of what he is not persuaded, or which arguments specifically he found unpersuasive, and exactly why they failed to persuade.

Perhaps it is the philosophical discussion which makes him so uncomfortable, because it forces the reader to make a choice between three incredible alternatives. The "aggressive tone" emphasizes the fact that this is a choice which cannot be conveniently avoided, because rejecting one alternative simply commits one to one of the remaining incredible alternatives. In my discussions with colleagues I have discovered that many feel very uneasy about having to make this choice, and would prefer to just avoid the issue altogether. The reviewer does not give us the benefit of his own thoughts as to whether it is Indirect Realism of which he remains unconvinced, or whether he is unconvinced of the validity of the argument which forces him to a choice in the first place. Unless he informs us otherwise, we can only guess at his motivations. But given the abundance of closet Naive Realists out there, the most likely explanation is that this reviewer is a Naive Realist at heart, his whole career has been committed to theories based implicitly on Naive Realist assumptions, and he feels uncomfortable at now being challenged to defend the indefensible. Like many psychologists this reviewer probably considers philosophical debates to be outside of his specialty, and therefore irrelevant to his branch of psychology. This is exactly why an "aggressive tone" is needed to wake these people up to the fact that psychology cannot be so neatly compartmentally insulated from philosophy, because every science is built upon a foundation of philosophical assumptions, and those assumptions are not always "testable" by the normal rules of evidence. Paradigmatic choices require the exercise of a kind of broad-minded judgement or intuition, which this reviewer probably considers to be unscientific.

And yet the choice must be made, because scientific theories which are built upon the wrong paradigmatic foundations are like castles built on sand, impervious to assault on the grounds of theory and evidence, but vulnerable to a blow below the paradigmatic belt. Given this reviewer's reluctance to share with us the reasons for his discomfiture, we can only guess what is motivations might be. But even if I have this reviewer's motivations all wrong, there are many more closet Naive Realists out there who are in need of an "aggressive tone" to wake them up to their responsibility as scientists to justify the foundations of their theoretical stance.


The author must provide a MORE STRUCTURED COMPARISON OF HIS PROPOSAL WITH OTHER THEORIES OF SPACE/FORM PERCEPTION. It is only against the background of existing proposals that his contribution can be assessed. And here I mean NOT GENERAL LINES OF INVESTIGATION, SUCH AS NEURAL NETWORK MODELS, BUT SPECIFIC *PERCEPTUAL* THEORIES.

Is it the volume of referenced material which the reviewer finds inadequate, or is the reviewer refering to some specific literature? Would the reviewer care to specify which "important findings" in the psychophysics of form perception he considers to be so relevant to the Gestalt Bubble Model? Or does he merely suspect that there may be some such findings and that the author should seek them out? But the real issue here involves the general nature of the proposal. The reviewer is unhappy that this paper is not a detailed and specific model that makes testable predictions which can be matched against experimental data. And this reviewer will not consider the present paper suitable for publication until it is revised accordingly. What he has failed to understand is the paradigmatic nature of what is being proposed. What I propose is not a detailed model of some specific perceptual effect, but a whole new class of model, motivated by the Indirect Realist perspective that perception involves the construction of a volumetric spatial replica of the external environment in an internal representation. What makes this a paradigmatic idea is the fact that if it should turn out to be right, then it would necessarily have implications on our interpretation of a large volume of psychophysical data. In fact, should this idea be proven right, virtually no aspect of psychology will be entirely unaffected by this new perspective on the problem. So the volume of psychophysical data which are relevant to the model would include much of the literature in psychology. Unfortunately that data cannot be definitive in determining the paradigmatic choice itself, because in paradigmatic debates, both sides often cite the same evidence to support their opposite conclusions. That is because each side interprets the data from the perspective of their own paradigm. For example the most convincing evidence for a spatial representation in the brain is the fact that consciousness exhibits a spatial structure. The structural nature of experience is an indisputable and undisputed fact. But that evidence is convincing only to the Indirect Realist. The Direct Realist interprets that self-same factual evidence as a property of the world rather than of the brain. The two paradigms draw opposite conclusions from the self-same evidence. And they also draw opposite conclusions from the evidence of mental imagery, neglect syndrome, visual illusions, Gestalt phenomena, and virtually every other domain of psychology.

The reviewer suggests a "more structured comparison" with other theories of space / form perception. "Not general lines of investigation such as neural network models, but specific perceptual theories". Here the reviewer clearly reveals his misunderstanding of the scope of the present proposal. For the competing hypotheses which the present model proposes to challenge are not specific perceptual theories, but exactly more general lines of investigation, such as neural network models. Models can be validly defined at many levels, from general concepts to specific mechanisms. Consider for example Marr's and Biederman's models of vision by abstraction of features; Selfridge's Pandemonium model; Triesman & Gelade's spotlight theory of attention; Collins & Quinlan's Spreading Activation model, McClelland & Rummelhart's PDP approach, to name just a few. Some of these models are far more general and conceptual than mine, but are valid models nonetheless. And then of course there is the example of Gestalt theory itself, a concept so general that it can hardly even be called a model. And yet despite its vagueness as to specific mechanism, the Gestalt view of perception serves as an invaluable reminder to resist the temptation to consider only simpler aspects of perception that can be described by specific models. It would have been a great loss if the Gestalt ideas had been denied publication because the concept was not sufficiently specified! General models are appropriate in a new field where much remains to be discovered, while more specific models are derived as more exact specification of general models, as a science matures. It is unfortunate that there are so many in the scientific community who consider general discussion of paradigmatic issues to be beyond the bounds of science, which they would reduce to a pedantic pursuit of minute details in hermetically insulated narrow specialty domains. That is a very impoverished view of the enterprise of science!

The set of PREDICTIONS PRESENTED SHOULD CONCERN PERCEPTUAL FACTS. After all this is a perceptual theory, not a theory of consciousness. For example, the last two predictions concern the nature of subjective experience, and are hardly predictions of the model.

Is it not the prerogative of the author to determine whether his paper is a theory of consciousness or a perceptual theory? What this reviewer fails to see is that every theory of perception is also a theory of consciousness, because the two are inseparable.

The "last two predictions" to which the reviewer objects are the fact that an illusory figure is experienced as a solid spatial surface at high resolution, and that the reversal of a multistable percept is vividly experienced as an inversion of a perceptual data structure. What he means is that I have not specified the computational algorithm of the model sufficient to perform computer simulations that reproduce those phenomena in detail. He is right, I have not. But the vast majority of models out there do not even consider it necessary for a model of the phenomenon to produce a volumetric spatial output. The novel and significant message of the present paper is to point out that an adequate model of the phenomenon must produce exactly such an output. This is the nature of paradigmatic hypotheses, they outline a general approach to a problem, not its detailed solution. It is only after the paradigm has been established that others will feel the need to construct those specific models and perform those simulations. If this reviewer had his way, the paradigm would never even be published as a theoretical possibility for other researchers to consider, so those models would never be built and the predictions never tested.

In summary, the paper has novel, interesting elements that have great potential to become a BBS paper. Nevertheless, the author needs to REORGANIZE THE PAPER SO AS TO HIGHLIGHT THE ACTUAL CONTRIBUTIONS IN A MORE STRUCTURED MANNER.

This is the kind of review that I hate most of all! In the first place the reviewer does not bother to summarize the paper, so we have no idea whether he read it or understood it to any depth, or more importantly, whether he understood the principal arguments as intended by the author.

Secondly his criticisms are so vague as to be essentially meaningless. He complains that the philosophical discussion is "not very persuasive", but he does not explain of what he is not persuaded, and why. He demands discussion of "important findings in the psychophysics of form perception" but does not specify what those findings might be. He says that the discussion of the literature should be "more comprehensive and in-depth", but he does not specify what issues should be discussed. He demands "more structured comparison" with other theories, but he does not specify which theories he considers relevant to this comparison. This kind of criticism is such a carte-blanche, there is no way that these vague requirements could ever be shown to have been met.

This reviewer appears to offer a qualified endorsement of the paper, i.e. revise & resubmit. But make no mistake about it, this is actually an outright rejection. Because the reviewer has made it clear that the kind of revision that he requires is the kind of revision which would transform the paper from a paradigmatic hypothesis to a simple perceptual theory. And as a perceptual theory it would be promptly rejected for publication because the volumetric filling-in operations would seem unnecessary and neurophysiologically implausible in the absence of the discussion of the paradigmatic issues. If this paper were revised to meet with this reviewer's satisfaction, I would no longer wish to be its author.