Surfaces and Essences

by Douglas Hofstadter

  In short, the non-military, non-historic image of dominos falling on a tabletop was profoundly mixed, in the minds of the leaders, with the political, military, and historical image of a set of countries that would fall, one after another, to invading forces (whether in Europe or in Southeast Asia). Given this, can one sharply distinguish between grand historical analogies and humble, mundane analogies that come from everyday life? Clearly not, since, as we’ve just seen, sometimes dominos on a table are seen as countries during a recent war, other times as countries during a current war, and yet other times as just plain old wooden dominos. (As Sigmund Freud might have said, “Sometimes a domino is just a domino.”) Anyone who knows the domino metaphor and the historical facts cannot help mixing all these images in an intimate fashion. (Later in this chapter we will carefully consider this kind of mental blending of situations. We use the term “frame blends”, while pioneering frame-blend researchers Gilles Fauconnier and Mark Turner opt for “conceptual integration networks”.)

  As we have already observed, most clearly in Chapter 3 and Chapter 4, any specific event — perhaps the falling of some dominos on a table, perhaps a famous historical defeat — can be encoded at various levels of abstraction, which means it is perfectly possible for us to see as one and the same phenomenon the succumbing of a series of neighboring countries to an evil empire’s domination and the successive toppling-over of many dominos neatly lined up in a chain. This universal fact of human high-level perception allows us to see far beyond the concrete details of situations and to connect events that superficially are enormously different from each other.

  We now come back to the notion that Khong proposed of a “non-analogical argument” based solely on abstract concepts. The problem is that whatever abstract concept is under discussion (dominos, containment, etc.) in a military or political context, it will necessarily evoke, simply because of the words it involves, familiar everyday images. Furthermore, in the mind of anyone who has a decent knowledge of history, an abstract concept of this sort will also evoke a wide range of historical analogues, at differing levels of awareness (for example, activation of a concept of this sort could evoke a range of historical precedents, such as a moat around a medieval castle, the walls of medieval cities, the Great Wall of China, the Maginot Line, and so forth). The fact that a word such as “containment” seems abstract and somewhat bland does not mean that it is devoid of a metaphorical substrate and that it evokes no historical precedent. To the contrary, the power of abstract words comes from the fact that they evoke a set of concrete images, all derived from experiences that one has had, either directly or vicariously, over the course of one’s life. This rich halo of familiar and historical analogues constitutes the imagery that is evoked in the mind of a decision-maker, and which in turn gives rise to life-and-death decisions.

  Pluralization and Schemas

  Khong’s book abounds in vivid expressions such as “No more Munichs!”, “another Hitler”, “a series of Koreas”, “the next Chamberlain”, and “replays of 1917”. As we saw in Chapter 1, when we discussed words like “Mother” and “Moon”, a one-member category is, from the moment that it comes to exist in a mind, capable of being pluralized and used just like any highly abstract notion. (Chapter 4, too, in the section dealing with Platos, popes, Mozarts, meccas, Bachs, and bibles, gave many examples of this sort.) In the above expressions (and others) in Khong’s book, a term like “Hitler” or “Munich”, seeming on its surface to denote a particular historical entity (a person, a place, a defeat, a war, a strategy) is used as the label of a general category, of which there could exist dozens or perhaps hundreds of instances, sometimes actual, and sometimes imaginary. As this makes clear, it is not even remotely possible to draw an ironclad distinction between “concrete” and “abstract” usages of a concept in one’s memory.

  Nonetheless, Khong tries to draw a clear line between analogues and so-called schemas. A schema is defined as the mental superposition of several historical precedents, whereas just one historical precedent all by itself would be an analogue (one side of an analogical bridge, to recall the image from Chapter 3). It’s tempting to imagine that this is a sharp and unambiguous distinction, but as we just saw, the concept of thirties, although Khong treats it as if it were one specific event, is not one event at all; he himself defines it by blurring together at least five different events, and so, by his own definition, it is a quintessential schema! And yet this plurality does not keep him from speaking about “the analogy to the thirties”, as if such a thing involved making a mapping to one single historical precedent, every bit as concrete, local, and specific as the Munich agreements or Mussolini’s annexation of Ethiopia. Thus we see that sometimes Khong says that a schema is an abstraction and that seeing an event as covered by a schema is not the making of an analogy, while other times he treats a schema and an analogue as completely indistinguishable (both serving as ends of analogical bridges).

  In Chapter 3, we saw that when someone reminds us of another person, what takes place inside our head is the building of an analogical bridge between two mental representations, and that when we categorize an object that we see, such as a cup, the same mental process is involved. In the case of the reminding, the mental entity that gets freshly activated is our memory of a specific person, whereas in the case of the categorization, the freshly activated mental entity is based on a lifelong series of perceptions of different objects (or situations), and we have little or no recollection of the “founding members” of the category. Despite this difference, the two processes are cut from the same cloth, since analogy-making is at their core; that is, in both cases, what is going on is an act of analogical mapping that builds a link between a fresh new mental representation and an older mental representation stored in our brain.

  Khong’s notion of a schema is not significantly different from a many-membered category such as cup, while his notion of a particular historical precedent is essentially the same as the mental representation we have of an old friend. One might think that mental processes involving schemas versus those involving historical precedents are very different, but all that is involved in either case is the building of analogical bridges between mental representations. Moreover, as we saw in the case of the slowly accreting Twain–Grieg–Einstein concept in Chapter 3, a schema can slide gradually from very concrete to very abstract, which means that it makes no sense to try to draw a sharp dividing line between making an analogy and using a schema.

  Let’s take a concrete example. If you are arriving for the first time in a lawyer’s office, you might have this anticipatory thought: “I’m probably going to have to wait for a long time in this waiting room, just as I did the last time I went to get a physical at Dr. Blahblah’s office.” Or you might instead think: “I’m probably going to have to wait for a long time in this waiting room, just like when I have doctor’s appointments.” Then again, you might well have this thought: “People who have private practices and fancy offices always make you wait a long time.” In the first case, which would seem comparable to one of Khong’s historical precedents, the source situation (the checkup with Dr. Blahblah) is equally concrete as the the target situation (the meeting with the lawyer), and you superimpose your recent experience of a long wait in your doctor’s office onto this new situation. In the other two cases, which would correspond to Khong’s schemas, the source situations are more abstract: a generic visit to your doctor, or an even more generic visit to any kind of professional. But in all three cases, you are using a familiar source analogue in order to make educated guesses about a brand-new situation. As soon as you scratch a little, you find that, under the skin, the putative distinction between using an analogue and using a schema vanishes into thin air.

  Domestic Politics and the Associated Brain Mechanisms

  In a section toward the end of his book, instead of pressing on with his major thesis that analogies are omnipresent in political thinking, Khong promotes the curious idea, pu
shed by certain historians, that analogical thinking plays no role at all in decisions about domestic politics. One can’t help wondering how the mental mechanisms involved in thinking about domestic politics could differ fundamentally from those involved in thinking about foreign events. Indeed, such a distinction is most implausible.

  Imagine a historian of physics suggesting that physicists, when tackling questions in thermodynamics, always depend on analogies, but that when tackling questions in electrodynamics, they never use them at all. Now why would one area of physics be resistant to a certain set of thought mechanisms while another area of physics required those exact mechanisms at all times? The idea is as silly as suggesting that Capricorns always use analogies but Geminis never use them. The fact is, as Chapter 8 will show, that all areas of physics depend on analogies. And what holds for thinking in physics should (by analogy!) hold just as much for political thinking. Indeed, there are good reasons for believing that the mechanisms underlying human thought are universal.

  Are We Humans Really So Superficial?

  If psychologists working on analogy were asked to name the most solidly established experimental finding in their discipline, surely the winner in such a poll would be the notion that surface-level features are the key to memory retrieval. In his book, Khong puts great stress on this empirical finding (and we could quote dozens of similar claims in articles describing experiments on analogy-making): “One of the most interesting findings of researchers working on analogical problem-solving is that people pick analogies on the basis of superficial similarities between the prospective analogue and the situation it is supposed to illuminate.”

  Experimental studies have indeed demonstrated that subjects who are shown a source situation and who are then given a target situation are usually unable to see any connection between the two unless they share surface-level traits. Furthermore, in such experiments, when two situations have a superficial resemblance, then the second one invariably brings the first one to mind, no matter whether it is appropriate or not (that is, irrespective of whether there are deeper reasons to connect the two cases). For instance, if subjects first tackle an arithmetic problem concerning items bought in a store, then any other problem concerning purchases will instantly remind them of the initial problem. But if the theme of the first problem is experimentally manipulated — say it becomes a visit to a doctor’s office instead of a store — then the participants will almost surely see no link between the two stories, even if the solution method for the first problem applies perfectly to the second problem.

  If such a broad claim were true — if in cognition the superficial always wins out — it would have profound consequences concerning both the quality of human thought and the utility of analogy as the basis thereof. If this claim were true, one would have to face the sad fact that we humans are simple-minded creatures able to react solely to the most obvious aspects of what we encounter, unable to see beyond façades constantly leading us astray, incapable of spotting deeper essences lying behind the scenes. In a word, it would mean that our brains would be like bulls constantly charging madly at beguiling red flags everywhere. The lure of the superficial would lead us to believe that everything that glitters is indeed gold, that seeing a swallow inevitably means that spring has sprung, and of course that books must always be judged by their covers. And analogy-making would come out looking pretty sorry, thanks to this law of the surface’s winning appeal, for it would be revealed as a primitive mode of thinking that relies on facile and misleading resemblances between things. This would really be grist for the mill for analogy’s numerous detractors. And from there it would be but a short step to conclude that we humans should promptly seek ways around this crude and unreliable method of thinking in favor of a more rigorous, more deductive method based on the mental manipulation of abstract symbols and guided by the precise time-tested laws of logic. Indeed, this would be the logical conclusion to draw!

  The cognitive psychologist Dedre Gentner, well known for her research on analogy, has proposed, with some colleagues, an evolutionary interpretation of the lure of the superficial (without necessarily subscribing to this interpretation):

  These findings may leave us feeling schizophrenic. How can the human mind, at times so elegant and rigorous, be limited to this primitive retrieval mechanism? An intriguing possibility is that in the evolution of cognition, retrieval from memory is an older process than inferential reasoning over symbolic structures. We could thus think of our surface bias in retrieval as a vestige of our evolutionary past, perhaps even a mistake in design that we have never lived down.

  If the ideas in this passage are valid, then the bias towards superficial analogy-making is a ball and chain that has shackled humanity since time immemorial and that continues to plague us to this day. Is our bias towards superficiality merely an unfortunate legacy bequeathed to us by a Homo sapiens but a Homo less sapiens than we are? If so, then hopefully all we need to do is wait a few hundred thousand years for natural selection to purge this bias from our systems.

  The domination of surface-level features in experiments on memory retrieval is a very robust phenomenon, confirmed by a great number of studies. There can be no doubt about the correctness of the finding. And yet to understand it properly — especially to understand why the evolutionary interpretation cannot be taken seriously — one has to go into the psychologists’ labs and get one’s hands dirty. The dirt one has to deal with, in this case, is the experimental paradigm that has guided nearly all experimental investigations of analogy-making, and which we’ll call the “source–target paradigm”. In this paradigm, subjects first study a source situation, which typically is a problem whose solution is given to them; then, at a later time, a new and unsolved target problem is presented to them and they try to solve it.

  What makes this paradigm so attractive to psychologists is how easily, following it, one can design experiments that can be performed quickly on a large number of participants. Furthermore, it is a powerful technique since, by varying the source of the analogy, one can compare the behavior of groups of participants each of which is exposed to a different source, or possibly to none at all. As a result, the source–target paradigm has totally dominated the world of experiments on analogy-making, because (alas!) there are some good reasons for using it, as we mentioned above. We say “alas” because this paradigm, much like a medication that is effective but that has serious side effects, has unfortunately helped to propagate misleading ideas about analogy-making.

  The Achilles’ heel of this paradigm — indeed, its fatal weakness — is that the analogies studied in experiments based on it have little to do with analogies made outside the laboratory, in “real life”. Both the humble analogies that we all make on a daily basis, as necessary to our survival as the air we breathe, and the flashes of genius that every so often light up the scientific landscape and give rise to a revolutionary new theory, are cut from a radically different cloth from the analogies that are generally studied in the laboratories. The community of researchers who investigate analogy-making experimentally has mistakenly extrapolated the results of their experiments following the limited source–target paradigm to the entirety of analogy-making. There is great irony here, since the researchers who assume this paradigm to be representative of all of analogy-making have themselves fallen for a misleading analogy.

  Unfortunately, the source–target paradigm has a serious defect that undermines the generality of the conclusions that experiments based upon it produce. This defect stems from the fact that the knowledge acquired about the source situation during the twenty minutes or so of a typical experiment is perforce very limited — often consisting merely in the application of a completely unfamiliar formula to a word problem. By contrast, when in real life we are faced with a new situation and have to decide what to do, the source situations we retrieve spontaneously and effortlessly from our memories are, in general, extremely familiar. We all depend implicitly on knowledge deeply roo
ted in our experiences over a lifetime, and this knowledge, which has been confirmed and reconfirmed over and over again, has also been generalized over time, allowing it to be carried over fluidly to all sorts of new situations. It is very rare that, in real life, we rely on an analogy to a situation with which we are barely familiar at all. To put it more colorfully, when it comes to understanding novel situations, we reach out to our family and our friends rather than to the first random passerby. But in the source–target paradigm, experimental subjects are required to reach out to a random passerby — namely, the one that was imposed on them as a source situation by the experimenter.

  And so, what do the results obtained in the framework of this paradigm really demonstrate? What they show is that when people learn something superficially, they wind up making superficial analogies to it. It would hardly be an earth-shaking revelation that people who have been given a single five-minute juggling lesson turn out to be lousy jugglers. Or suppose that subjects were taught the rules of chess in two minutes and then were made to play a few games. Could one validly draw the conclusion “human beings employ very primitive strategies when they play chess”? Of course not. And yet this is the character of the “scientific conclusion” that superficiality trumps depth whenever people make analogies. Are we poor human beings really so constantly gulled by surface appearances?