Constraint on the semantic flexibility in visual statistical learning

Abstract

We examined whether the expression of visual statistical learning (VSL) is flexible at the superordinate-categorical level. In the familiarization phase, participants viewed a sequence of line drawings. In the test phase, participants observed two test sequences (statistically related triplets versus unrelated foils) that consisted of the same objects as those presented during the familiarization phase (the same condition) or different objects that shared the same categorical information with drawings during the familiarization (the different condition). They then decided whether the first or the second sequence was more familiar. The results of Experiment 1 showed greater familiarity above chance levels for statistically related triplets only in the same condition. In Experiment 2, even where the word stimuli representing each superordinate-level category were included in the test phase (the categorical condition), the results showed VSL only in the same condition. Our findings suggest that the semantic flexibility of VSL is limited to the basic-level category.

Keywords:

• Visual statistical learning
• Flexibility
• Object recognition
• Categorical information

This work was supported by JSPS KAKENHI [Grant Numbers 11J02236, 13J08125].

Research in this article was previously presented at the 53rd annual meeting of the Psychonomic Society and the 77th annual convention of the Japanese Psychological Association.

This work was supported by JSPS KAKENHI [Grant Numbers 11J02236, 13J08125].

Research in this article was previously presented at the 53rd annual meeting of the Psychonomic Society and the 77th annual convention of the Japanese Psychological Association.

Notes

¹ In line with Turk-Browne et al. (2008), participants were asked whether they noticed if the triplet had any structure during the familiarization phase after the test phase. In fact, participants were asked: “Did you notice any patterns in the visual stream of line drawings in the first session (i.e., the familiarization phase)?” and then “If so, please recall these patterns and write the name of the line drawings in serial order.” None of the 40 participants reported all of the triplet structure in the familiarization phase across our experiments. Almost all participants did not report awareness of statistical regularities or stated that they noticed only one triplet, or one or two pairs of objects that consistently appeared together. These reports do not necessarily mean that VSL measured with the 2IFC familiarity judgement was based on implicit process. Turk-Browne, Jungé, and Scholl (2005) discussed that “the familiarity judgment is an odd sort of dependent measure to use, because it essentially asks observers to make an explicit judgment about implicitly learned relationships”. Recently, some studies examined this issue with a rapid visual serial presentation (RSVP) task as an implicit measure of VSL (Bertels, Franco, & Destrebecqz, 2012; Kim, Seitz, Feenstra, & Shams, 2009). Thus, we should use an implicit measure of VSL, instead of explicit familiarity judgement, to examine exactly whether our results were based on implicit process or not. However, this problem is out of the scope of this paper, and we will address this point in future research.

² It is possible that an explicit measure like familiarity judgement in the test phase is not sensitive to extracting the semantic flexibility of VSL even if the participants could have slightly learned the regularities at the superordinate level in the familiarization phase. Thus, we will need to examine the same issue in future research with, for example, an RSVP task described in Footnote 1. Nevertheless, it is important that we demonstrated the lack of semantic flexibility at the superordinate level with 2IFC familiarity judgement used in most studies of VSL.

³ We thank the editor and anonymous reviewer for suggesting this point.

⁴ We thank an anonymous reviewer for suggesting this point.