Vision and Inference

Charles D. Gilbert, "The Constructive Nature of Visual Processing," in Eric R. Kandel, James H. Schwartz, Thomas M. Jessel, Steven A. Siegelbaum & A. J. Hudspeth, eds., Principles of Neural Science 556, 556-557 (5^th ed., 2013):

Vision is often incorrectly compared to the operations of a camera. Unlike a camera, however, the visual system is able to create a three-dimensional representation of the world from the two-dimensional images on the retina. In addition, an object is perceived as the same under strikingly different visual conditions.

A camera reproduces point-by-point the light intensities in one plane of the visual field. The brain, in contrast, parses scenes into distinct components, separating foreground from background, to determine which light stimuli belong to one object and which to others. In doing so it uses previously learned rules about the structure of the world. In analyzing the incoming scream of visual signals the brain guesses at the scene presented to the eyes based on past experience.

This constructive nature of visual perception has only recently been fully appreciated. Earlier thinking about sensory perception was greatly influenced by the British empiricism philosophers, notably John Locke, David Hume, and George Berkeley, who thought of perception as an atomistic process in which simple sensory elements, such as color, shape, and brightness, were assembled in an additive way, component by component. The modern view that perception is an active and creative process that involves more than just the information provided by the retina has its roots in the philosophy of Immanuel Kent and was developed in detail in the early 20th century by the German psychologists Max Wertheimer, Kurt Koffka, and Wolfgang Koehler, who founded the school of Gestalt psychology.

The German term Gestalt means configuration or form. The central idea of the Gestalt psychologists is that what we see about a stimulus--the perceptual interpretation we make of any visual object--depends not just on the properties of the stimulus but also on its context, on other features in the visual field. The Gestalt psychologists argued that the visual system processes sensory information about the shape, color, distance, and movement of objects according to computational rules inherent in the system. The brain has a way of looking at the world, a set of expectations that the rides in part from experience and in part from built-in neural wiring.

&&& The dynamic evidence page

Evidence marshaling software MarshalPlan

Brain Endowments and the Interpretation of Natural Signs

We have known for some time that bits of evidence function as signs, or clues. (Sherlock Holmes knew this.) But we do not have a good sense of what it is in us or in the cosmos that enables us to take little bits of evidence and extract from them, possibilities, conjectures, and, even, conclusions (inferences) that reach far beyond the evidence that seems to inspire or provoke them. Now we may have some further support for a conjecture about part of the answer. The source of that support? Pigeons.

Birds are excellent navigators. Indeed, they are phenomenal navigators. But scientists have not been able to explain very well how birds, with their teeny-weeny brains, manage their navigational feats of derring-do. It has been suspected that the answer lies in the ability of birds to detect earth's magnetic fields. But no one has been able to demonstrate the part of the anatomy of birds that enables them to detect and interpret magnetic fields. Until now. See James Gorman, Study Sheds Light on How Pigeons Navigate by Magnetic Field NYTimes (April 26, 2012). It appears that there are "cells in a pigeon’s brain that record detailed information on the earth’s magnetic field." Id.
Stunning. So is there also reason to believe that human brains are also soft-wired with devices that facilitate the interpretation of tiny "evidentiary clues," sensory signals, or "natural signs"? It must be so, yes? The conscious mind -- explicit human thought -- is not by itself capable of extracting all the things (some of them "true," or, in any event, indicative or suggestive of true propositions) that human beings, like pigeons, manage to extract from the environment.
This may not be "intelligent design." But it is, in a way, a kind of harmony of the spheres, a harmony between the internal mechanisms of organisms -- such as pigeons and human beings -- and their "environment" (the cosmos). The only really peculiar thing about the human animal is that it is able, sometimes, to also deploy conscious thought -- explicit ratiocination -- to interpret eeny-weeny signs, or hints(?), in its environment.
Or so I speculate.

• Cf. the second part of my essay Are There Universal Principles or Forms of Evidential Inference? Of Inference Networks and Onto-Epistemology in J. Jackson, M. Langer & P. Tillers, eds., Crime, Procedure, and Evidence in a Comparative and International Context (Hart Publishing, 2008) & William Twining, Philip Dawid & Dimitra Vasilaki, eds., Evidence, Inference and Enquiry (Oxford & British Academy, 2011).

The mighty pigeon...

&&&

The dynamic evidence page
Evidence marshaling software MarshalPlan.

Of Inference Networks and -- Gasp! -- Onto-Epistemology

&&&
The dynamic evidence page
Evidence marshaling software MarshalPlan
It's here: the law of evidence on Spindle Law. See also this post and this post.

Causation, Explanation, and (what shall we call it?) Quasi-Intuition, a wee bit of Tacit Knowledge, or a touch of Unknowing Knowing

One possible escape from the riddle of causation and inference that I have been exploring is the notion that there is a distinction between causal theory and explanation: there are (some observers think) valid or good explanations that are not yet -- or theories or accounts that fall short of -- potentially deeper causal theories or explanations.

If this distinction between causal and other explanations works, it provides a solution to the mystery of why "mere association" bereft of causal theory works. The suggested answer: sometimes mere association is not mere association at all: the so-called mere association sometimes rests on an explanation.

But in science -- and perhaps (one hopes) in other fields as well -- mere explanation is not enough, an explanation must be a good explanation).

In science -- and, again one hopes, in other intellectual domains as well -- a theory, even a "merely explanatory" one [n.b., my phrase], must prove its mettle, it must be put to the test. In science this often means that it must be shown that the theory or explanation in question can predict the occurrence or non-occurrence of phenomena or events better than pot-shot strategies do, better than helter-skelter guessing does.

By embracing "explanation" some theorists do seem to be advancing the claim that a theory short of a causal one can achieve have predictive force, that it can achieve a genuine understanding of Nature without having in hand anything that we might call a theory of what causes Nature to act as it does.

But if explanation sans causality can achieve such understanding -- if it can indeed have this kind of epistemological power --, a big question quite naturally arises:

How? How does "explanation without causation" achieve such understanding?

The difficulty here is that a "mere explanation" falls short (by hypothesis) of a causal theory: it is short of, it is less than, a theory that rests on an understanding of the mechanisms or principles that (may) underlie phenomena and events in nature.

The move to "explanation" suggests that those who make this move accept the notion that ignorant human beings can (somehow) achieve (scientific) understanding. Paolo Garbolino -- revealingly and perceptively -- points to this issue (and attempts to resolve it) by suggesting that the proper epistemological alternative to reliance on (i) causal explanation and, alternatively, (ii) statistical explanations is the deployment of (iii) "potential explanatory accounts" [P. Garbolino, "Explaining Relevance," The Dynamics of Judicial Proof: Computation, Logic, and Common Sense 179, at 187-191 (M. MacCrimmon & P. Tillers, eds., Physica-Verlag, 2002].

A problem is not solved merely because it is named -- even if the name happens to be a seductive word such as "explanation." (But, note, attaching a name to a problem or process sometimes does advance understanding "simply" by identifying a problem, phenomenon, or process).

The label "explanation" does not fully explain -- but whoever thought it would? --, the label "explanation" does not resolve the question of how ignorant human beings -- human beings who lack a full understanding of Nature and the mechanisms or principles that drive or explain it --, a phrase such as "explanatory theory" does not by itself explain how such ignorant human beings sometimes manage to make very good judgments about the behavior of Nature.

The riddle just posed may provide us with part of an answer; it may at least yield or suggest a useful "non-answer answer," viz., a partial answer that helps even if it does not altogether satisfy the hunger for knowledge.

The non-answer answer I have in mind is that human beings know more than they know, they have understanding that eludes their comprehension, they have some tacit knowledge.

There can no longer be any serious doubt that this is the case; it is no longer possible to doubt that people have knowledge that they cannot articulate, spell out, make (fully) explicit. (This general insight is not new; it goes back at least to Plato.)

But, of course, there is at least one big problem with being told that you know more than you know, that you already have much knowledge: sometimes -- and now is one of those times! -- you would like to know more than you now know. Does it do any good to be told you know more than you know? For example, does the notion of unknowing knowing tell us -- or help us to decide -- how a judge should instruct a jury to think about evidence that suggests that this or that substance causes cancer or does or when if ever evidence that this or that substance causes cancer or did cause cancer should be withheld from a jury?

So -- to repeat the general question -- is the notion of unknowing knowing helpful?

Possibly.

I must, alas, suspend discussion of this big question for now. But let us leave our conversation (are we having a conversation?)--, let us suspend this discussion with one caveat in mind: no theory of knowing that fails to account for advances in human knowledge can satisfy. It is true that know more than we know. But we also manage to learn; there are new things -- and new insights -- under the sun; and knowledge -- including our knowledge of how we know -- is not entirely circular.