To the Editor:

Kevin Shapiro’s extended review of my book, The Emotion Machine, was excellent in most respects, but he apparently misunderstood the aim of the book’s form [“How the Mind Works,” May]. He writes:

Like Minsky’s other popular books, The Emotion Machine has its own peculiar style of exposition, relying heavily on neologisms (like “imprimer”), diagrams, epigraphs (sometimes repeated), and imagined interlocutors (identified by labels like “Student,” “Psychologist,” “Romanticist,” “Determinist,” and so forth). Despite its straightforward and relatively unadorned prose style, it can be difficult to follow—in part because the ideas it contains are not presented or organized in any particularly sensible way. (Not the best advertisement for critics, selectors, and encouragers in action.) Its pages are peppered with references to Minsky’s own prior writings and what amount to literary rain checks: promises that an issue will be discussed at some later point.

A recent email I received reflects a different point of view: “I love the style of writing wherein you inject opinions of Citizens, Students, etc. To my mind, your words flow cleanly and clearly, and I hope that I can find an editor for my book who will make my writing flow as wonderfully.”

Here is one hypothesis to account for our critics’ divergent views: some readers find that to understand a complex idea, a diagram is worth a thousand words; others prefer a few sentences to albums of graphic images. My “peculiar style of exposition” aims to reach both such audiences, as well as general readers and advanced researchers.

Mr. Shapiro suggests that The Emotion Machine is a popularization of earlier works of mine, but in fact most of its theories have not been published before. They also do not quite lend themselves to being “organized in a sensible way”—such as a conventional narrative style—because the central idea is based on the following theme:

If you “understand” something in only one way, then you scarcely understand it at all—because when something goes wrong, you’ll have no place to go. But if you represent something in several ways, then when one method fails, you can switch to another. That way, you can turn things around in your mind to see them from different points of view—until you find one that works for you!

Accordingly, I argue that our brains have evolved in multiple ways to serve our most important mental functions. This suggests that psychologists should not always try to emulate the technique that works so well in other branches of research: namely, to seek the simplest explanations. Because a human brain must not only execute different methods but also select which among them to deploy and when, it was my task to describe the methods themselves, how they might be acquired, how they are controlled, and how the brain picks up the slack when they fail.

Naturally, a writer cannot do all of these things at once, and must rely on Mr. Shapiro’s “literary rain checks”—but the promises are carried out in due course. Hundreds of students helped me find ways to organize the ideas, but some of them needed time to “cook” before others could be built on top of them. (Perhaps I should add that my book takes some time to understand; its contents occupy a full-semester course.)

Let me add a few specific remarks.

• Mr. Shapiro complains that my “discussion of imprimers covers ground trod more than a half-century ago by the school of object-relations psychologists.” This is true, but my chapter about “attachments and goals” covers new ground by adding new and constructive ideas about how Freud’s “introjection” process might actually work to build, inside a child’s mind, representations of the values and goals of the persons to whom he becomes “attached.”

• To my statement that “I don’t know of any experiments to see if structures [for linking the same things seen from different points of view] can be found in our brains,” Mr. Shapiro responds that “scores of such experiments have been conducted over the last few decades.” But here I was discussing a new idea—“Parallel Analogy” or “Panology”—about the structures of human memories. Designing experiments to look for these structures would be difficult, because they would depend on neural connections between analogous fragments of knowledge that would not be easily seen with the current crop of instruments.

• Finally, I was startled by Mr. Shapiro’s statement that “[a]s for the larger mysteries—like the nature of meaning and the origins of consciousness—it is not clear that Minsky has much to say beyond identifying the solutions he thinks are bad.” This makes me wonder if he took the time to read carefully the two long and very detailed chapters titled “Consciousness” and “Commonsense.”

Marvin Minsky
Massachusetts Institute of Technology
Cambridge, Massachusetts

_____________

Kevin Shapiro writes:

I should begin by saying that, as a great admirer of Marvin Minsky’s work, I am honored that he took the time to respond to my article. His main bone of contention, it seems, is essentially a matter of style. Mr. Minsky meant for his book to be all things to all readers, which would be a difficult proposition under the best of circumstances—and even more so, one would think, when it includes the contents of a graduate-level course in computer science.

If this was the goal, it was probably inevitable that Mr. Minsky’s meshwork of new ideas would sometimes be hard to follow, however “cleanly and clearly” his words flow at the level of the paragraph or subsection (and on this point I quite agree with his e-mail correspondent). Perhaps it is fairest to say that The Emotion Machine is a book that is best appreciated when re-read.

Mr. Minsky seems to have missed my point about the structures he proposes to call “Parallel Analogies.” As he surely appreciates, there is a difference between asking whether a particular kind of representation exists in the brain and whether we can “see” the neural underpinnings of that representation using currently available instruments. I agree that we are very far away from understanding the neurophysiology of any kind of cognitive representation. But it is eminently possible to discern whether (and how) different kinds of knowledge are linked in the mind—for example, by studying how access to knowledge breaks down with injuries to the brain.

Needless to say, I did read Mr. Minsky’s chapters on “Consciousness” and “Commonsense,” and I hope he will forgive me if I found them less than satisfying. His main argument about consciousness is that it is not a clearly defined concept—that is, “consciousness” refers to any number of distinct mental processes, into which we have varying degrees of insight. Somehow, he suggests, all of these processes contribute to the “mental work” of choosing goals and courses of action, in a way that does not require a single “self” or central executive. This is certainly plausible, but it is not a new idea, and it does not explain exactly how people come to have (and act from) an integrated model of subjective experience. On this key point Mr. Minsky offers only more promissory notes. Although he seems optimistic that they will be redeemable in due course, I am somewhat more skeptical.