Straat Science

I returned to the Katholieke Universiteit Leuven in September 2004, about 14 years after a first, much briefer stay there. Visa requirements involved a fairly rigorous background check. I had to be fingerprinted, for example, in order to get my FBI rap sheet, a mandatory part of the visa application. When I finally arrived in Belgium, I had to apply for a national identity card, which required several trips to an office near the statehouse, as well as a police visit to my apartment to make sure I was who I claimed to be.

By coincidence, one of the projects I was pursuing in Leuven had to do with the way people trace identity. We often have to re-identify an object from a glimpse or a description—for example, to figure out that the rented car we’ve just spotted in the parking lot is the same car we drove earlier in the day. Perceptual studies of object recognition offer part of the story of how this is possible, but we must sometimes re-identify objects without the benefit of perception. The person we meet at our thirtieth high school reunion may have only the faintest resemblance to the classmate we last saw at graduation, and in these situations we have to figure out identity from non-perceptual, historical clues. So the question is, how do our concepts of individual people and things support these inferences?

Straat Sign: Emiel Vliegenberghstraat has become Maria Van Belstraat, but remains on the sign for reference.

My first few weeks in Leuven underscored the difficulties of identity, even apart from the waits at the government office for my identity card. I was living in the same compound and working in the same department as in 1990; but, of course, much about the city had changed, and my memory of places and people was vague. Was this the same café I had liked back then, a newly redecorated descendant, or a completely different place? Was the woman attending a seminar the same person as the graduate student I had met many years before? (Indeed, it was. Belgian academics stay very close to the nest.)

I spent many hours walking around Leuven since I didn’t have a car, and it eventually seemed as if even the city’s layout posed puzzles about identity. The map of part of the city shown here highlights a fairly typical intersection, not far from where I was living, and the accompanying photograph is of a street sign at this single intersection. Heading northeast on Constantin Meunierstraat, you find that the street divides three ways, the middle branch being Meunierstraat again and the flankers Maria Van Belstraat and Emile Mathieustraat. It’s reasonable that the central branch should be identical to the trunk (as in an ordinary, right-angled intersection), but this means that adjacent houses have abruptly changing addresses. Van Belstraat has an identity crisis of its own. The stripe through the upper street sign in the second photo indicates that this one-block street was formerly known as Emiel Vlieberghstraat. History is not erased here, but simply lined out. (The fine print on the street signs also provides historical detail: Van Bel was a miniaturist, Vliebergh a Flemish jurist. None of this information, however, will help you find your way through the maze of streets.)

From the perspective of identity problems, the issues were: What determines that one thing is the continuation of another (Meunierstraat) rather than an entirely new object (Mathieustraat or Van Belstraat) that first appears at the transition point? What dictates that one thing (Van Belstraat) comes into existence while another (Vlieberghstraat) departs?

In some ways, the street changes are the spatial versions of some traditional problems about identity in time. The most famous of these puzzles, elaborated by Thomas Hobbes, is about a legendary wooden ship, the Ship of Theseus, which was repaired over a long period by replacing its old planks with new ones until it consisted entirely of new planks. Hobbes’ twist to this story is that an antique collector hoards the old planks until he has the complete set and then reassembles the planks into an exact replica of the ship. Which is Theseus’ ship at this later point, Old Planks or New Planks?

I was working on a cognitive theory of object identity, in collaboration with Sergey Blok and George Newman, which we thought might address some of these problems. According to this theory (based on an earlier idea by Robert Nozick), people will decide that an object at a later time is identical to an object at an earlier time if two conditions are met. First, the later stage must be a causal extension of the earlier one. The person we meet at our thirtieth reunion, for example, had better be a causal outgrowth of our classmate at graduation or else she is simply an impostor. This causal relation must also be fairly close. If you run a magazine through a shredder, the excelsior that emerges causally grows out of the magazine, but probably isn’t close enough to qualify as the original. Second, the identical item must be the closest of all the causally close-enough contenders. The Ship of Theseus story is a puzzle because there are two objects that are causally close enough to be the original. Which ship is genuine may depend on how we weigh the opposing factors of having the same planks versus having a continuous functional connection with the initial ship.

Straat Ahead?: Rips points out that the streets of Leuven often find themselves in an identity crisis.

The key to the theory is that identity judgments depend on a double comparison. A later object is identical to an earlier one if it is close enough to be a contender and also closer than the rest of its close-enough competitors. Blok, Newman, and I were able to fit a quantitative version of this double-comparison process to the results of experiments in which participants read stories about object transformations and then made judgments about whether the outcome was identical to the original. For example, one experiment described an iceberg that split into two pieces (the size of the pieces varied across trials), and participants determined which of the two pieces (if either) was the original.

Does the same idea apply to spatial identity for streets and other routes? The names of streets are partly at the whim of the city government, which can re-christen Vlieberghstraat as Van Belstraat if it sees fit to honor a painter over a legal scholar. And although there may be a causal connection between a street and its extension in the street’s planning or construction, the trunk of a street doesn’t currently “cause” its branches. Nevertheless, the same sort of double comparison that we applied to more typical objects might also work with streets. Suppose you’re walking through Leuven at night following Meunierstraat and want to stay on the same street to find your destination. If you come to the intersection and can’t read the street signs, which branch do you choose? You might first decide whether any of the branches are close enough to dead ahead to qualify as Meunierstraat. If more than one are close enough, you then need to select among the winners. This means that a branch can be disqualified as the continuation in two ways: if it’s not close enough (for example, not near enough to dead ahead) and if it’s dominated by a better competitor.

To find out about this, Blok, Newman, and I conducted a lab version of the street identity task in which participants saw Y-shaped diagrams that we described as maps of routes (streets or rivers). The angles and widths of the branches (top spokes of the Y) varied across trials, and we asked participants to decide whether the left-most branch, the right-most branch, or neither branch was the continuation of the trunk (bottom spoke). The results of the study produced both effects that the model predicted: a minimum effect due to the closer of the two branches and a difference effect due to the separation between them.

I finally received my identity card—formally, my Certificate of Inscription in the Aliens Register—a few weeks before I left Belgium. By then I had also learned to find my way around the city without too many wrong turns, mostly by studying maps of the town like the one shown here. Identity cards and maps, though, are just the records of complicated threads of individuals. What the work in Leuven brought home—and what I brought home with me—is the thought that, despite a lot of cognitive research on exemplar theories and an enormous amount of research on recognition, we know hardly anything about the recognition of exemplars. Gregory Murphy and Douglas Medin famously pointed out that there must be cognitive glue that unites objects into the categories to which they belong. Much the same must be true of the individual objects themselves. Something must glue the discrete glimpses we have of these things into the individuals of which they are part. For typical natural objects and artifacts the cognitive glue is likely to be causality, as Blok, Newman, and I argue. Other sorts of objects, such as conscious creatures like us or winding streets of medieval cities, may use other types of identity glue, but there may be a process—a form of historical comparison—that unites these ways of uniting things.

For more on this topic:

• Blok, S., Newman, G., & Rips, L. J. (2005). Individuals and their concepts. In W-k. Ahn, R. L. Goldstone, B. C. Love, A. B. Markman, & P. Wolff (Eds.), Categorization inside and outside the lab (pp. 127-149). Washington, D.C.: American Psychological Association.
• Murphy, G. L., & Medin, D. L. (1985). The role of theories in conceptual coherence. Psychological Review, 92, 289-316.
• Nozick, R. (1981). Philosophical explanations. Cambridge, MA: Harvard University Press.
• Rips, L. J., Blok, S., & Newman, G. (in press). Tracing the identity of objects. Psychological Review. Washington, D.C.: American Psychological Association.

The Fulbright Scholar Program is sponsored by the Bureau of Educational and Cultural Affairs of the US Department of State and administered by the Council for International Exchange of Scholars.

Fulbright grans offer a variety of educational activities, primarily university lecturing, advanced research, graduate study, and teaching in elementary and secondary schools. The program has had more that 250,000 participants since its inception.

For more information, on the Fulbright program, visit www.cies.org or http://exchanges.state.gov/education/fulbright.