New Books from Daniel Dennett and Douglas Hofstadter

Daniel Dennett's new book is Intuition Pumps and Other Tools for Thinking and Douglas Hofstadter's new book (written with French psychologist Emmanuel Sander) is Surfaces and Essences: Analogy as the Fuel and Fire of Thinking. Eric Banks from Bookforum reviews the two new books.

This Is Your Brain, On

Two books seek to explain how our minds work their way through the maze of consciousness

ERIC BANKS





YOU DON’T HAVE TO CONDUCT A THOUGHT EXPERIMENT to see why some philosophers or scientists want to write for an audience cheerfully indifferent to the ways of the seminar room and the strictures of the refereed journal. Beyond the fame and fortune, perhaps more important is the sense that if one’s work is worth doing at all, it ought to reach the widest possible audience, particularly when it bears on issues (religion, free will) with decisive implications for how readers choose to live. Some, I imagine, also relish the bonus frisson of mixing it up in the rowdy rough-and-tumble of the public arena. If you’re like Daniel C. Dennett—one of whose many mantras is Gore Vidal’s “It is not enough to succeed. Others must fail”—what’s the point of felling a philosophical tree if there’s no one to hear it? Since the publication in 1991 of his book Consciousness Explained, Dennett has gladly risen to the challenge, merrily taking on all comers left and right, in works that play to a packed house most philosophers couldn’t dream of.

For Dennett, moreover, the experience of communicating to a broad readership his brawny materialist agenda, which aims at nothing less than squaring philosophy with a host of other fields—cognitive psychology, brain science, evolutionary biology—has an ancillary and less obvious boon. Specialists, he writes, tend to underexplain to one another the very terms of their discussions. These experts benefit from translating their respective positions down, as it were, so that they might be presented to “curious nonexperts,” as Dennett puts it in his newest book, Intuition Pumps and Other Tools for Thinking. They will be forced to think anew, and paradoxically to think harder: “To explain their position under these conditions helps them find better ways of making their points than they had ever found before.”

The notion that an idea or “position” might get fine-tuned just as neatly in the imagined company of a well-intentioned fast learner as it would among scholarly peers is ingrained in Dennett’s go-go style of doing philosophy and its winner-take-all stakes. As set out in Intuition Pumps, his narrative approach, with its flurry of catchy neologisms, plain-talk prose, and gotcha argument stoppers, will prove as roundly appealing to some as it will seem pandering, I suppose, to others. The pep-talk jocularity and the shoot-from-the-hip posture of its presentation, familiar enough throughout Dennett’s writing, make it seem as if the book—which is focused less on a single subject than on a kind of survey of Dennett’s greatest hits (from positions on consciousness to free will, from “intentional stances” to “competence without comprehension”)—imagines its reader as a slightly nerdy college kid with high math SATs who, with just the right writerly nudge, might be tempted to jump majors. There’s no accounting for tastes, but the odd recipe Dennett produces here—one part avuncular guide (who dubs himself “Uncle Dan” early on), one part pugnacious tough guy—makes for a weird slaw. Picture a helpful Burl Ives crossed with a philosophical Robert Conrad from the old commercials for Eveready batteries, just taunting anybody to go ahead, try and knock this position off—I dare you.

Dennett declares that his aim in Intuition Pumps is to lay out devices by which we might think more clearly, or with more insight, about a host of thorny topics—which might be boiled down to those many areas in which we errantly or too hastily assume we have a solid sense of the right and wrong answers. The sheer number of these thought experiments, geared to reveal how thoroughly incorrect our assumptions might be, is daring itself. The most provocative comprise consciousness, free will, and our own sense of what we mean by meaning and intend by speaking of intentionality—in other words, the philosophical terrain Dennett has explored extensively in his prior books. (A glance at the notes reveals how vastly Intuition Pumps recycles material and arguments he has used or made in previously published work, extending as far back as 1969.)

The Karnak Temple, Thebes, Egypt, and the Grand Canyon, Arizona.

Part of Dennett’s role in Intuition Pumps is to serve as a kind of design engineer. With the concept of “intuition pump,” he repurposes the thought experiment—a form of argumentation of ancient and venerable purpose in philosophy (and in sundry other disciplines, especially physics)—in order to transform its somewhat neutral-sounding disposition into a power tool, one that answers to a basic question: Is it well or poorly designed to get the job done? (The interesting question of the epistemological standing of thought experiments in philosophy is never really given much attention here.) First rechristened as “intuition pumps” in The Mind’s I, the hybrid work Dennett co-produced in 1981 with his friend Douglas Hofstadter, these narrative devices can condense, in a straightforward way, a complex set of propositions and suppositions into an imaginable story that summarizes or illustrates a position. Hence their extreme popularity in the history of philosophy, from Plato’s cave to Parfit’s amoeba. They can be positive or critical, launching a new idea or yanking the rug out from under someone else’s pet position (or even both). Either way, such thought experiments are designed to jolt the hearer’s or reader’s sense of intuition (hence the idea of the “pump” that paradoxically strands the reader’s analogic mind in an awkward, ill-specified locution) and channel it in certain indubitable directions. To mix the metaphor, intuition pumps are thus double-edged: They can accomplish a lot, and produce a kind of free analogue of the costly experiments carried out in laboratories, but they can also carry a heavy cost for a thinker when they are dubiously or even dangerously built. The lesson that Dennett hopes the reader might take away is that we must remain wary of intuition pumps, taking them apart to find their hidden biases and built-in assumptions, before we let ourselves be overinflated by them.

I lost track of just how many intuition pumps Dennett ticks off in the book, but taken as a whole they offer a sense that philosophy is a fabulous field if your career in sci-fi fails to take off. A menagerie of mindless robots, devious neurosurgeons implanting replica brain cells in their unknowing subjects, parallel worlds with exactly a single element changed from ours, swamp men transformed by strikes of lightning into brain-duplicating doppelgängers: Almost all these personae and philosophical fables will be familiar to anyone who has followed the rough course of the philosophy of mind for the past several decades. Some of the signal thought experiments devised by philosophers since the early 1970s are present: Mary the color scientist, who emerges from a black-and-white world; John Searle’s Chinese Room experiment, the response he devised more than thirty years ago to challenge an argument made in 1950 by artificial-intelligence pioneer Alan Turing. If a computer could pass for a human being to an interlocutor who wasn’t aware he or she was communicating with a computer, Turing held, the machine could be said to possess intelligence. As familiar as Searle’s Chinese Room experiment may be, it is worth lingering on it, since it illustrates what is ultimately so frustrating about Intuition Pumps. To quote Searle’s summary of the experiment:

Imagine a native English speaker who knows no Chinese locked in a room full of boxes of Chinese symbols (a data base) together with a book of instructions for manipulating the symbols (the program). Imagine that people outside the room send in other Chinese symbols which, unknown to the person in the room, are questions in Chinese (the input). And imagine that by following the instructions in the program the man in the room is able to pass out Chinese symbols which are correct answers to the questions (the output). The program enables the person in the room to pass the Turing Test for understanding Chinese but he does not understand a word of Chinese.

Dennett famously objected in The Mind’s I, as he does here, that Searle’s experiment pumps the wrong intuitions—among other things, it massively misrepresents what it means to “manipulate the symbols” or to “understand.” The example is undercooked and unnourishing, and if Searle had known anything about computing, he couldn’t in good faith have constructed it as he did. Therefore it is a flawed intuition pump: “It persuades by clouding our imagination,” Dennett writes, “not exploiting it well.”

But what is the difference between a good intuition pump and a flawed one? The Chinese Room has spawned scores of counter–thought experiments, replicating itself in many variations of structure and content; by the mid-’90s, Steven Pinker commented that it had become the source of at least a hundred papers. It has allowed articulations of positions from a vast number of academic fields, from proponents of AI to linguists, and generated commentary on semantics and syntax, intentionality and consciousness, and evolution. Sounds like a pretty fecund little tool for thinking to me! But for the budding philosophy student reading Intuition Pumps, Dennett reserves the right to select the hammer and pick the gauge of nail. “Here I am concentrating on the thinking tool itself, not the theories and propositions it was aimed at,” Dennett hopefully backtracks. But what good is it to present this book as a collection of helpful “tools for thinking” when it turns out the only successful tools happen to run on precisely the same voltage as Dennett’s own particular theories and propositions?

Intuition Pumps is valuable in providing an overview of a body of recent work in the philosophy of mind, but it suffers as well from Dennett’s penchant for cleverness—no more egregiously than in his soi-disant playfulness in mapping nasty flaws on his favorite intellectual targets, like Stephen Jay Gould. It grows tiresome and tacky: He returns to a long-ago pissing match with Gould to discuss rhetorical sleights of hand, and even coins a new word to describe the tendency to advance straw-man arguments and false dichotomies—“Goulding.” How is that a better “thinking tool”? He mocks philosopher Ned Block’s use of the word surely as a sure sign of a mental block (get it?) and offers up “Occam’s Broom” as an example of an argument that sweeps inconvenient facts under the rug, and circularly, not to mention condescendingly, takes the opportunity to chide Thomas Nagel for not consulting “the experts” on evolutionary biology. (At least he doesn’t call this oversight a “Nageling.”) All this sour score-settling with Dennett’s philosophical peers is infinitely less witty than I imagine he takes it to be. But in the spirit of Dennett’s tactic, I’d offer one historical vignette that characterizes his frequent summoning of an army of scientists at his back, with an arsenal of cutting-edge knowledge about our chemistry and biology, ready at some later date to vindicate his positions, and call that future-perfect feint a Ledru-Rollin. That would be in honor of the hectoring French propagandist of 1848 who famously bellowed, “There go my people. I must follow them, for I am their leader.”

Intuition Pumps at least has the benefit of tasking us with thinking about what we do when we intuit a given set of problems. Working with and against intuitions is a strategy that also permeates Surfaces and Essences: Analogy as the Fuel and Fire of Thinking, a book-length thought experiment published by French psychologist Emmanuel Sander and Dennett’s old partner, Douglas Hofstadter, probably still best known for his 1979 Pulitzer Prize–winning gift to high-SAT-math kids everywhere, Gödel, Escher, Bach: An Eternal Golden Braid. The new book brings together a laundry list of often laborious found examples, culled by the authors from their daily experiences and overheard talk, in order to tease out the logical paradoxes and contradictions of analogy, which, they contend, forms something like the essence of cognition. If that sounds like a sweeping claim, it is because in their view—as robustly flushed out in Surfaces and Essences—the nature of analogical thinking is vastly more complex than our folk understanding of it. The book’s argument may be stated fairly simply: When we as subjects attempt to make sense of any phenomenal experience in the world, which we do at every waking moment, we do so through a kind of quick cognitive shorthand, forging analogies between the unknown and past experience, both consciously and unconsciously.

The magic of analogical thinking is its odd recursiveness, a plasticity that has long delighted Hofstadter and engaged his fascination with metalanguage and the brain-teasing conundrums of self-referential puzzles. Analogy as a rhetorical device is an almost endless source of such entanglements. What makes one analogy like another analogy? They’re both analogical. Their definition as analogies refers to, well, other analogies. From this angle, which is very much Hofstadter and Sander’s preferred vantage, there may be naive analogies, and there may be analogies that lead us into dangerous directions or that cause us to make poor judgments, but it’s hard to see what would count as a wrong analogy. As such, in Hofstadter and Sander’s view, analogies exemplify a form of mental mapping, tying various states of things together or announcing semiresemblances among different types of experiences as we pass through the world. This schema links analogical thought profoundly with perception. Analogies can be beacons of creative thought, but they can also be utterly banal—just as our perception can be at times. And as much as we try to control our ability to dazzle and amuse with new analogies, it’s more frequently the case that analogies have a hold over us—over our language, over our thought patterns.

Here’s a sort of example. I was recently trying to describe to someone the (to my mind) unusual fact of my dog—of a breed very highly marked as “American” (a coonhound)—spending several long weekends without me in rural France. What suddenly popped into my head was the analogy of Tom Ripley, played by Dennis Hopper, in Wim Wenders’s The American Friend (1977), a classic of the New German Cinema with an American actor occupying an unexpectedly European landscape. Where did that come from? But my mind did another turn, as odd perhaps as the first, and I blurted out the film as Fassbinder’s The American Soldier (1970), another “Americanism” of German film. A collision of analogies! It took some time for me to figure out the train of “A is to B” at play, but the important point is threefold. First, the mental linking, which felt unmotivated, no matter how “creative” the thought, was so rapid that it felt automatic. Second, I had this set of analogies primed by an experience that seemed to call forth a pseudocategory—experimental German films of the ’70s that, according to the logic of experimental German films of the ’70s, featured an American component (whether in their titles or their casting)—which it’s hard to imagine I might have stored somewhere as a useful “category,” years ago, just waiting for an analogical item to happen onto the scene. Finally, the multiple, linked frameworks involved (German films with Americans in them, films that have American in the title) are flexible enough that they could blend into one another to create in essence a makeshift, almost ad hoc new frame. (The downside of such conscious awareness on my part is that it’s hard not to look at the poor pooch now and not think of Dennis Hopper.)

Hofstadter and Sander’s book is a bottomless exploration of the potential of analogic thinking to eat away at any simple idea of how one thing is related to another. The authors pursue this problem by pondering analogies of all types and at ascending levels of abstraction, with lists that span pages. They posit that the logic of analogy brings together not just how two unrelated things get related by their likeness (A is to B as X is to Y), but how, on a different level of abstraction, we might analogize relationality itself. We make sense of a variety of situations, sometimes consciously, sometimes below the level of reflection, by thinking of them in terms of label-like proverbs or aphorisms or even fables: X situation is just like what we think of when we think of the experience of “sour grapes,” or of “have your cake and eat it too.” The abstraction that provides the “label” for an analogy may not even have a name: Hofstadter returns throughout the book to a recurring and evidently personally haunting example of a situation in which the act of gathering stray bottle caps while on a tour of the temple at Karnak is linked to a memory of his young son, during a family visit to the Grand Canyon, mesmerized by a formation of ants instead of the sublime view. There’s no convenient or pithy category label for what these two things share, though the manner in which Hofstadter processed the former experience, as he exhaustively argues, seems to depend on an ingrained version of analogical thinking.

This being a book with Douglas Hofstadter as an author, it will discursively scale the Karnak–Grand Canyon experience in the form of a poem written by a friend, plumb the loop-de-loop relationship between the analogy thus formed and the incipient category it instantiates, and construct a tower out of analogies that ensue from yet another . . . analogy. It will show how the ur-form of a particular analogy (“X is the Y of something”) will throw off endless variations (in one virtuoso list, Hofstadter and Sander offer found variations including the “Bill Gates of wastewater,” the “Tiger Woods of user-generated video,” and the “Mussolini of mulligatawny”). Surfaces and Essences is a Hofstadterian machine of knot tying: With lists after lists, some virtuosic, some groan inspiring, of how to do things with analogy, it becomes clear that analogy and categorization are inseparable. We use analogies whenever we open our mouths—it should be obvious in the sentence above that “scale,” “plumb,” and “tower” are just the tip of the iceberg, so to speak, of how pernicious analogical language is on some basic level of communication—and Hofstadter and Sander eat up 590 densely printed pages thick with puns to make sure that we don’t miss the point. Surfaces and Essences caps itself with a twenty-five-page-long dialogue between two characters, Katy and Anna, who, like a pair of escapees from a play by Brecht, debate the positions taken in the book, with the dummkopf Katy arguing the bad view that categories form the core of cognition while the enlightened, analogically hip Anna proves the errors of her ways. (Wouldn’t you know it—every argument that Katy comes up with for her view about the primacy of categories as the basis for analogy turns on conceptualizing “categories” by way of various analogies. Like a spin on Monsieur Jourdain, she didn’t know she was speaking in analogies all along!)

Rhetorical strategies aside—insert metajoke here—you have to wonder whom this book was written for. It seems like a textbook, but there are no notes, no bibliography, no real sense of how Surfaces and Essences fits with or argues against (analogous?) work in psychology or cognitive science. It’s difficult, too, to gauge its urgency. The book feels like it could have been hatched three decades ago, around the time that George Lakoff and Mark Johnson wrote Metaphors We Live By, which raises many similar issues. Maybe that’s irrelevant, and the point is to disabuse us of a conceptual, context-free model of “dictionary meaning”—that meaning is a matter of discrete taxonomic categories like “mammal” or “sandwich” or “president,” all of which will by definition contain a set of necessary and sufficient features. Fair enough (although who really believes that?). But when all thought becomes “analogized” as analogy, if it, like Bertrand Russell’s tale of the turtles, is analogy all the way down, the explanatory value of “analogy” comes to seem tautological. This may be the way cognition “works,” but if there’s no other way to express it but, well, analogically—then it frankly seems a game of increasingly clever wordplay. To take a further step and refer to “analogy” as the core, that is, essence, of thought is to cast the analogy as hard fact, which seems no less reductive than the kind of conceptual models from which Hofstadter and Sander have labored mightily to rid us. The analogy I’m thinking of is “can’t have it both ways.”


~ Eric Banks is the former editor in chief of Bookforum and the former president of the National Book Critics Circle.