Presidential Lecture Series banner
Douglas Hofstadter
Stanford Humanities Center

Douglas R. Hofstadter:

Analogy as Core, Core as Analogy


Douglas R. Hofstadter
© Douglas R. Hofstadter. Used by permission.


Douglas R. Hofstadter, a noted author, cognitive scientist and proud Stanford son, returns to Stanford in February, 2006, with a lecture called “Analogy as the Core of Cognition.” Hofstadter is College of Arts and Sciences Professor of Cognitive Science, and Director of the Center for Research on Concepts and Cognition, at Indiana University. Although he is the author of eight major books, Hofstadter is best known for his magisterial first work, Gödel, Escher, Bach: an Eternal Golden Braid (a.k.a. GEB:EGB [a.k.a. GEB]), published in 1979 and awarded the Pulitzer Prize in 1980. In a sense, this dazzling and well-known work is the core of all of Hofstadter’s subsequent writings: it certainly contains the seeds of the ideas and formal approaches that fill those later writings — the core being the natural place to look for seeds, after all.

Or perhaps the throwaway apple core is not the best analogy for Hofstadter’s thoroughly delicious first book. Perhaps a better analogy is the “core” of “core curriculum,” for GEB has certainly become a centerpiece in the education of many students of artificial intelligence, intelligent artifice, the mind, and the “I” — not to mention the more traditional fields of computer science, psychology, linguistics, et al. There is a generation of readers who came of age when GEB was fresh and new and long on the New York Times bestseller lists, and for whom GEB may have been the first “serious” book they read: for this generation, GEB was a core part of a largely extracurricular general education — a work that taught us, on the cusp of the latest Information Age, that we ourselves could bridge the gap between C.P. Snow’s “two cultures,” between left-brain and right-brain concerns and approaches, and that our new computers (along with new approaches to mathematics, and music, and art) might help us to do so. Indeed, by encompassing and synthesizing aspects of these very diverse fields Hofstadter has helped us to understand that perhaps what is most important in human cognition takes place in some holistic center, some core of our mind, rather than on the “right” or the “left” side of our crania.

Unless by “core” we mean something else entirely: for example, the sort of core that a nuclear reactor has, the place in which all the action occurs, the fissions and chain reactions and transmutations and creation of new elements. In a revealing preface to the Twentieth-Anniversary Edition of GEB, Hofstadter divulges not only that his first book is not really about Gödel or Escher or Bach, popular opinion (and its title!) notwithstanding — but also that this popular opinion mystifies him somewhat: “Many people think the title tells it all: a book about a mathematician, an artist, and a musician. But the most casual look will show that... [t]here’s no way the book is about these three people!” Still, although this opinion is clearly better than the brief “description” given in the newspapers (“A scientist argues that reality is a system of interconnected braids.”), Hofstadter is never one to encourage misunderstandings or unintended readings, and in this same Twentieth-Anniversary Edition preface, he offers among “The Key Images and Ideas that Lie at the Core [!] of GEB” this very clear and concise expression of his topic: “how it is that animate beings come out of inanimate matter.” In many ways, this topic (along with its many manifestations and transformations) could be said to be at the core of many of Hofstadter’s writings.

Gödel, Escher, Bach is a lively weave of what we think of (or at least thought of, pre-Hofstadter) as rather disparate thematic threads. But these thematic threads are only the warp of Hofstadter’s tapestry: its woof is the book’s set of lively and remarkable formal features, on both micro- and macro-levels. Of course the book has all the structural attributes one would expect in a book: chapters, lists, tables, notes and indexes, for example. But between chapters are curious Carrollian dialogues that serve the obvious purpose of introducing theoretical (mathematical, formal, etc.) topics that will be treated in a (somewhat) more traditional manner in each upcoming chapter. These may also provide what might be called comic relief from some of the intensity of propositional calculus and discussions of formal systems — but it would be short-sighted indeed to see these as merely entertainments that allow us to step outside the formal constraints of the serious mathematical book: so far, so Gödel. But no, these dialogues are much more: as they comment on and embody characteristics of musical forms they become a Bachian Bacchanal of formal play, and a serious attempt not only to illuminate one art with another art’s light, but also to translate between and among art forms. Likewise, as Hofstadter’s formal play is inspired by a very serious purpose, so is his formal seriousness bedecked with playfulness: you’ll find in the bibliography a citation for an imaginative (and imaginary) work of Hofstadter’s alter ego, Egbert B. Gebstadter, and in the index an entry leading purposefully (and ever so playfully) to a trio of typos in the main text — all together a real pot-pourri.

Le Ton beau de Marot: In Praise of the Music of Language, from nearly twenty years later (1997), is also a book of both theory and practice: its principal topic is translation (in the most traditional sense), and it is, of course, replete with practical examples — including eighty-eight translations of a single brief poem by Clément Marot, “A une Damoyselle malade,” sometimes known by its first line, “Ma mignonne” — the almost minimalist form of which (28 three-syllable lines) makes Hofstadter’s massive treatment seem all the grander. Clearly, Hofstadter’s principal interest is not Marot at all (a lesson we should have learned from GEB), but rather poetry translation generally — and perhaps less translation for translation’s sake, than translation as a means to understand language, and language as a means to understand human thought. This is not to say that the actual translations play second fiddle to the book’s “cognitive science” basso ostinato: on the contrary, the multitude of translations of Marot, along with close readings and discussion of them, are the theme-and-variations vehicle through which Hofstadter conveys his most profound ideas — or, once again, the formal woof and thematic warp.

Although GEB continues to be widely seen as Hofstadter’s “trademark” work, he expresses some disappointment that Le Ton beau — in his introductory words to the book (words which he will reaffirm a few years later), “probably the best book I will ever write” — did not seem to take its place in the hearts of many of his readers. Hofstadter’s own explanation for this — and his decision to include it in so prominent a place in GEB’s own Twentieth-Anniversary Edition preface — are telling: “In some sense, GEB was a ‘forward-looking’ book, or at least on its surface it gave that appearance [...] Well, by contrast, Le Ton beau de Marot might be seen as a ‘backward-looking’ book, not so much because it was inspired by a sixteenth-century poem [...] but because there simply is nothing in the book’s pages that could be confused with glib technological glitz and surreal futuristic promises. Not that GEB had those either, but many people seemed to see something vaguely along those lines in it....” Perhaps what these readers mistook for “glib technological glitz” in GEB is actually something they would also find in abundance in Le Ton beau de Marot: Hofstadter's characteristic charisma.

Structurally and formally, Le Ton beau de Marot is as playful and odd as was GEB. In a place analogous to GEB’s inter-chapter dialogues, Le Ton beau’s chapters are separated (or joined?) by poetic interludes of translation and discussion, the pages of which are numbered separately to enhance the sense that this is a distinct structural part of the book. But Hofstadter’s remarkably formalist and intellectual book about poetic translation is, at the same time, rich with emotion: in addition to its occasional joyous bawdiness (in a jocular translation and discussion of a Marot paean to the breast) and some no-holds-barred polemics (against Nabokov in particular, which one critic unfairly called a “tantrum” — it’s not, although it is an emotional and heartfelt critique), Le Ton beau is deeply, remarkably, and perhaps unexpectedly moving. Without a hint of the maudlin, the book relates the story of the heart-rending death of Carol Hofstadter, the author’s still very young wife. But the story of this personal tragedy has much more than a bit part in Le Ton beau: in fact, Carol’s presence and absence, and Hofstadter’s abiding love for her, are tangible throughout, very literally from cover to cover: from the multiple puns in the title, through the dedication, the introduction, the translations, all the way to the endnotes and index... and even in the front-cover art and dust-jacket author photo, which has a hidden meaning relating to Carol. (By the way, speaking of endnotes and index — here, as everywhere else in Hofstadter’s oeuvre, phantom works, hidden gems and oddly moving moments grace not only the text proper, but also the entire apparatus; it pays to pay close attention.)

Recurring throughout Hofstadter’s work is this question of animacy arising from inanimacy: thought and language and feelings blooming from unthinking, unspeaking and unfeeling stuff; beauty and depth arising from places inherently outside the realm of the aesthetic. Gödel, Escher and Bach are clearly representative of the animacy/inanimacy question — but so, for example, is Hofstadter’s own love for his wife, which endures even across the boundary between life and death: can the beauty and depth of all these things really be expressed fully in mere marks on a page, transmitted as mere sound waves or as patterns of light and dark? And on an even deeper level, how are these marks and patterns created, and how is their resulting depth and beauty perceived, in the physical matter of the brain? Hofstadter certainly addresses the “mind-brain problem” explicitly to pose the question in one particular and narrow form. But he is interested in much more than just this narrow question: he is interested in “meta-acts” which themselves embody or even transcend the question (hence his choice of three particularly “meta” characters in GEB). In many cases, though, the seemingly mundane act of translation becomes a transcendent meta-act for Hofstadter. The persistent awkwardness of modern machine translation belies the opinion that translation is a mundane and mechanical act, and is especially striking given the relatively advanced state of natural language mimicry.

Hofstadter’s continuing interest and activity in translation touches not only (and not even principally) Marot — although Le Ton Beau de Marot is by far his fullest treatment of the “problems ”of translation — but also of his own work. As he tells the story of this topic, “A very different strand of my intellectual life was my deep involvement in the translation of GEB into various languages, and this led me, perhaps inevitably, in retrospect, to the territory of verse translation” (GEB, p.P-20). It is the “inevitably” that seems the most astute, for now it seems that Hofstadter’s work, at its “core,” was always about translation, in any of a large number of its multifarious guises. Still, Le Ton beau does focus on a specific type of formal and highly structured and constrained translation of the most traditional sort, between human languages. And Hofstadter characteristically imposes even greater constraints, by focusing on a single poem, itself with remarkable formal constraints. Even within these tight strictures, though, an entirely new work emerges: a sizable chunk of Le Ton beau is dedicated to the problems of translating Alexander Pushkin’s Eugene Onegin, arguably the most important work in Russian literature.

Onegin not only offered Hofstadter a test case for comparative translation in Le Ton beau, where he dedicated several chapters to it: it also became his next major work, this time as a translator rather than author (although his “Translator’s Preface” to the work is a major essay in itself).Onegin is not only a much larger canvas than was “Ma Mignonne” (5300 lines, in nearly 400 tightly-structured sonnet-like stanzas) on which to practice his art of poetic translation. Onegin is also an ideal object on which Hofstadter can focus his artistic and intellectual interests: as its own “formal system” (a central concept in GEB) it is highly disciplined; at the same time, it is a linguistically playful, sparkling work. And yet, through all its formalism, within all its formal constraints, it remains a deeply romantic and moving novel of love and loss. All of these characteristics lie at the core of Hofstadter’s own artistic and intellectual value system — in Pushkin he had clearly found a kindred spirit — and his translation, completed for the bicentennial of Pushkin’s birth (1999), both reveals and revels in their shared passion for language, precision, play and spark, in ways that most previous attempts have not. Hofstadter’s “novel versification” (as he wittily subtitled it) of Pushkin’s famous “novel in verse” is both a loving tribute to the original and a wonderful addition to the family of Onegin translations into English.

Fluid Concepts and Creative Analogies (1995) is a set of essays and experiments of Hofstadter and a sort of cognitive science circle (fittingly called the “Fluid Analogies Research Group,” part of the aforementioned Center for Research on Concepts and Cognition) that he has led at Indiana University for many years. This collection reveals in concrete, programmatic form many of the facets of Hofstadter’s own concerns as a thinker and writer in cognitive science: efforts in “artificial intelligence” (a term and an approach that Hofstadter at first embraced, then later traded in for the more comprehensive — and more fitting — name, “cognitive science”), creativity studies, computer modeling of human cognition, and, of course, analogy. Analogy here is presented not only as the “core of cognition,” but also as a core concern for any computer modeling of human thought. And here it becomes clear that Hofstadter’s own primary concern in “artificial intelligence” is not to create a machine that thinks, but rather to use a machine to model, and thus better understand, how we think. As Hofstadter writes in his closing words to Fluid Concepts, he’s aiming not for programs that better mimic natural human language, for example (which he calls mere “window dressing”), but rather for computer “architectures that go much further... toward capturing the genuine fluid mentality” of the way we think — that same fluidity for which Alan Turing famously hoped to test his black box in what became known as the Turing Test.

Shortly after GEB’s smashing success, Hofstadter and philosopher Daniel Dennett “composed and arranged” and, in 1981, published The Mind’s I: Fantasies and Reflections on Self and Soul, a marvelous collection of essays and short fiction, each piece followed by a “reflection” by one of the editors. In some ways the structure of this book seems more standard, and perhaps less experimental, than Hofstadter’s other books. But the collection itself, as an anthology, is a marvel: Borges fantasies appear alongside academic works on artificial intelligence, programming, psychology, philosophy, paradoxes and puns. (Pre-dating and pre-dicting Fluid Concepts’ resistance to mistaking flashy natural-language “AI” programs for true human-like cognition, Hofstadter here waxes eloquently ludic: “Is a soul greater than the hum of its parts?” [p. 191]) It is a sign of his continuing curiosity and generosity that he sees so much value in such a wide-ranging set of works, engages with them so deeply, and shares both the works and his engagement with his readers in such an approachable collection. Such a mix of genres, such enthusiastic engagement — even such a meaning-packed pun! — is typical Hofstadter.

The other major work of “early Hofstadter” is Metamagical Themas: Questing for the Essence of Mind and Pattern, a collection of his Scientific American columns written in the early 1980s, along with several previously unpublished essays. (Hofstadter’s column in Scientific American succeeded the venerable and long-running column of Martin Gardner called “Mathematical Games.” His curious title is, characteristically, an anagrammatic homage to Gardner’s.) Here too Hofstadter’s authorial generosity and enthusiasm shine: since much of the work comes from a monthly column, the reader normally “implied” is here made flesh, as Hofstadter quotes liberally from letters and contributions he’s received from Scientific American readers. Metamagical Themas is very much in the spirit of GEB, and even includes some sections that had been excluded from it, such as the marvelous — and marvelously neological — Tortoise/Achilles dialogue “Who Shoves Whom Around Inside the Careenium?” In a typically Hofstadterian strange-loopy twist, in this dialogue Hofstadter appears not only as author, but also as a character in the dialogue of his own fictional creations. Here not only is the “implied reader” fully realized — the “real author” is fully implied, too.

Ambigrammi: un microcosmo ideale per lo studio della creatività (1987) is a curious, and, sadly, little-known work, published only in Italian. Its principal topic is one of several droll and quirky art forms that pepper (and otherwise spice up) Hofstadter’s larger works: ambigrams (a word of Hofstadter’s own coining), that is, text forms that are symmetrical, in that they can be read in several directions (for example, forward and backward, in rotation, in mirror image, etc. — in fact, in sixteen distinct symmetrical ways). Some of Hofstadter’s other “cognitive” art forms, such as his musically-inspired “Whirly Art” and his letter-inspired “gridfonts,” were shown in a recent exhibition; Hofstadter discusses all these forms in the exhibition catalog — and they do all receive some explanation and treatment throughout his larger works — but ambigrams remain by far the most completely illustrated, explained and elaborated, in both the examples and taxonomy contained in the book, and in a fanciful dialogue between Hofstadter and his erstwhile alter ego. The ambigram form encapsulates many aspects of human cognition, and Hofstadter explicates and exemplifies these here much as he does those aspects of poetry translation, formal logical systems, musical forms, etc., in his other books: perception and categorization, formal constraints, elasticity and fluidity of thought, creativity, play, and analogy — and though he takes each of his many art forms quite seriously, it’s clear that he is insterested not only in each particular what of human creativity and cognition, but especially in the how.

During his career Hofstadter has written, lectured, experimented and, above all, thought about an exceptionally wide variety of human endeavors, all in pursuit of that question he has posed continually since writing GEB more than a quarter century ago: “how it is that animate beings come out of inanimate matter.” After this incredibly varied quarter century of work, is it possible to concede that the answer to this question lies in the simple statement that he forcefully makes in the title of his Stanford Presidential Lecture: that analogy is the very core of cognition?

True, there have been plenty of other candidates in Hofstadter’s work for the “core” of “cognition” — cognition itself being the “core” of the animacy that is contained in our fundamentally inanimate careenia, the cogito of our sum, the I of our mind. Is the core of cognition and animacy essentially only self-representation and self-reference (as in Bach, in our DNA and elsewhere)? Is it essential incompleteness (as in Gödel’s Theorem and elsewhere)? Is it strange loops and tangled hierarchies (as in Escher, in Hofstadter’s own forthcoming book, I Am a Strange Loop, and elsewhere)? Is it in the patterns, puzzles, paradoxes, puns, poetry, and programming that we see throughout Hofstadter’s work? Or is it elsewhere?

Elsewhere.... Perhaps it is precisely in analogy that we find the common thread of all these cognitive and creative phenomena, and thus the common element in the endeavors that make us human, and thus the core of our humanity. We await the answer — or at least a fascinating exploration of more facets of the question — at the core of this lecture: it will be here, if it’s anywhere at all.

Really, though, is this the last word in Hofstadter? Of course... not. Who could even imagine such a thing?

Text by Glen Worthey
Humanities Digital Information Service

Stanford University Libraries

Thanks to Ever Rodriguez for Web site
and other technical assistance.

©2006 Stanford University Libraries


Top of Page || Home Page || Stanford University Libraries || Stanford University