When we count objects, counting is relatively fast with small sets, i.e. less than four objects. This fast enumeration of objects is called subitizing. Several theories were proposed to explain this phenomenon. In this work we propose that the arrangement of objects is critical in the speed of enumeration of objects, and there is only one model that can account for this feature: the pattern recognition theory. According to this theory, if we can see only a few objects they can be perceived as a pattern (yeah, sounds simple), and recognizing the pattern is enough to recall the appropriate number, e.g. triangle is three, or square is four, etc.
In this specific work participants enumerated objects. The objects were arranged either in a symmetrical and well known configurations (the so called canonical patterns) or in a random pattern. Additionally, we used two further arrangements: slightly and strongly distorted canonical patterns. Participants could subitize sets with more than four objects if they were arranged canonically. Furthermore, the more noisy the patterns were, the less likely the participants could subitize them. These results are consistent with the ideas that (1) subitizing is pattern recognition, and (2) the pattern recognition system is flexible enough to transfer any pattern of less than five objects to a known pattern, but beyond four this ability is limited.
We gave further theoretical considerations in our paper:
- The pattern recognition model is mostly rejected in the literature because usually a modified version of it is cited instead of the original version (see the introduction of our paper).
- Many former results are consistent with the pattern sensitivity feature, and consequently support the pattern recognition model of subitizing, even if this aspect is not explicit in some of those works (see introduction and conclusion parts of our paper).
- The limitation of four nicely fits to the limitation in similarity detection of the recognition system (see the conclusion of our paper).
- Finally, a more comprehensive model of subitizing is proposed based on our result, the former results, and the theoretical considerations (see the conclusion of our paper).
Krajcsi, A., Szabó, E., & Mórocz, I. Á. (2013). Subitizing Is Sensitive to the Arrangement of Objects. Experimental Psychology, 60(4), 227–234. https://doi.org/10.1027/1618-3169/a000191
Find a PsychoPy script of the experiment.
See a short slideshow about some of the ideas below. Note that the graphs do not necessarily show the final published data.
Description of the project in Hungarian.