Projects: Concepts

Most research on concepts in psychology has centered on concepts of categories like people or trees or tables. Concepts of individual items — a particular person or tree or table — have mostly been taken for granted, perhaps because psychologists regard them as given by perception. It’s clear, though, that we’re able to follow individuals over long stretches of time, during which we may not have perceptual contact with them. How do we keep track of their identity over time?

Rips, Blok, & Newman (2006) and Blok, Rips, & Newman  (2005) propose a causal theory of object identity, based on an earlier proposal by Robert Nozick. To decide which object at a later time is the same object as an earlier one, people determine which is the causal outgrowth of the original item. Rips et al. and Blok et al. provide some evidence for this theory, based on participants’ judgments about the outcome of object transformations. They also outline a formal (mathematical) model that predicts these decisions. We have extended this research to look at questions about pronoun reference (e.g., can the causal connections expressed in a passage resolve the intended antecedent of an ambiguous pronoun? See Sagi & Rips, 2014). We’ve also looked at paradoxes involving intransitive judgments of identity (where object A is identified with B and B with C, but where object A is judged not identical with C); Rips  (2011). A related project with Dan Bartels deals with the way a person’s identity over time affects his or her decision to consume a resource now versus delay gratification; Bartels & Rips (2010).

As a contrast with concepts of physical objects, research with Susan Hespos has examined infants’ knowledge of nonsolid substances like water and sand; Hespos, Ferry, & Rips, 2009;  Hespos, Dora, Rips, & Christie, 2012; Hespos, Ferry, Anderson, Hollenbeck, & Rips, 2016. Rips & Hespos (2015) review theories of the distinction between object and substance concepts and advance a new theory of this difference. Rips & Hespos (2019) look at the way languages realize this distinction. Research on this topic dovetails with work on number (see the research page on reasoning and math), since we tend to count objects but measure substances.

Other work on concepts concerns the way notions of part, kind, and boundary apply across the domains of objects and events. Certain verbs (e.g., eating) have boundary (or aspectual) conditions that depend on the object noun with which they are paired. Eating a chocolate bar is an event that concludes when the bar is finished, but eating chocolate sauce is an event that is potentially unbounded. Other verbs (e.g., watching, delivering) don’t depend on the boundedness (count or mass status) of their objects. We’re looking at analogies between bounded and unbounded verb types and bounded and unbounded noun types; Wellwood, Hespos, & Rips (2017).

Some earlier papers on natural categories and essentialism appear in the Publications page of this web site.