Transitive inference is an ability to derive a relation “Mary is taller than Kate” from the premises “Mary is taller than Ann” and “Ann is taller than Kate”. A typical nonverbal transitive inference task involves a series of overlapping simultaneous discriminations, for example A+ B-, B+ C-, C+ D-, and D+ E-(where letters stand for arbitrary stimuli and pluses and minuses denote reinforcement and nonreinforcement). In the test, a subject is presented with a critical test pair BD, in which both stimuli were reinforced in one pair and not reinforced in another pair. A choice of the stimulus B over the stimulus D is interpreted as indicative of transitive inference (if B > C and C > D, then B > D, therefore choose B).

​Accounting for an effect of reinforcement history on the choice of stimulus B over D in a critical BD pair has been one of the most significant challenges in a history of nonverbal transitive inference research. Broadly, the choice of B over D could be interpreted as an indicator of transitive inference, or as a result of richer reinforcement history (i.e., higher probability of reinforcement over time) for a stimulus B than for a stimulus D. As a part of my dissertation (done at Moscow State University), I have developed a procedure to control for potential influences of reinforcement history. Using this procedure, I have shown that crows behave in a transitive-like manner (i.e., select stimulus B over D) despite the opposite predictions by reinforcement-based models of TI (Lazareva et al., 2004). 

​At the University of Iowa, I was able to replicate these results using pigeons (Lazareva & Wasserman, 2006) and adult humans (Lazareva & Wasserman, 2010).I have also explored whether it is possible to predict choice behavior in the transitive inference test by measuring associative strengths of the individual training stimuli, as reinforcement-based models of transitive inference suggest. My results demonstrate a lack of strong relationship between empirical measure of associative strength and pigeons' choice behavior (Lazareva & Wasserman, 2012), questioning basic premises of reinforcement-based models do not appear to be supported by empirical data.

​Currently, we are conducting extensive modeling of TI data obtained in different species. Clara Bergene has worked on modeling rhesus monkeys' data obtained at Emory University by Robert Hampton and Regina Paxton Gazes, and is currently preparing these results for publication. David Goodman has completed the TrI toolbox that will allow other researchers to evaluate predictions of reinforcement-based models for their data. 

We have also explored the involvement of hippocampus in transitive inference in pigeons. Research with mammals (both rodents and primates) suggests that transitive behavior is dependent on hippocampus. In contrast, a single prior study using pigeons reported no effect of hippocampal lesion on TI (Strasser et al., 2004). However, this study used a very unusual behavioral technique that could influence the results. In our study, we used traditional TI technique plus a control for influences of reinforcement history. This project has been started during Summer 2011 by Kaitlyn Kandray who received DUSCI summer award to conduct this research in collaboration with Martin Acerbo (Iowa State University). The results of this project suggested that hippocampal lesion impairs transitive-like behavior in pigeons that do not rely on associative values during test; pigeons that are using these values are not affected by hippocampal lesions (Kandray, Acerbo, & Lazareva, 2015). 

Published papers:

Lazareva, O. F., Kandray, K., & Acerbo, M. J. (2015). Hippocampal lesion and transitive inference: Dissociation of inference-based and reinforcement-based strategies in pigeons. Hippocampus, 25, 219-226.

Paxton Gazes, R., Lazareva, O. F., Bergene, C. N., & Hampton, R. R. (2014). Effects of spatial training on transitive inference performance in humans and rhesus monkeys.  Journal of Experimental Psychology: Animal Learning and Cognition, 40, 477-489.

​​Lazareva, O. F., & Wasserman, E. A. (2012). Using resistance-to-extinction and resistance-to-reinforcement to measure associative values of stimuli B and D in a transitive inference task. Behavioural Processes, 89, 244-255.

Lazareva, O. F., & Wasserman, E. A. (2010). Nonverbal transitive inference:  Effects of task and awareness on performance. Behavioural Processes, 83, 99-112.

Lazareva, O. F., & Wasserman, E. A. (2006). Effect of stimulus orderability and reinforcement history on transitive responding in pigeons. Behavioural Processes, 72, 161-172.

Lazareva, O. F., Bagozkaja, M. S., Smirnova, A. A., Zorina, Z. A., Rayevsky V. V., & Wasserman, E. A. (2004). Transitive responding in hooded crows requires linearly-ordered stimuli. Journal of the Experimental Analysis of Behavior, 82, 1-19.

Nonverbal transitive inference

Comparative Perception and Attention Laboratory