In order to focus on the understanding of specific problems, he identified three levels (ordered according to decreasing abstraction):Ĭomputational level (problems that cognitive ability has to overcome) Īlgorithmic level (the algorithms that may be used to achieve a solution) and Marr (1983) proposed a commonly accepted general framework for analyzing levels of explanation in cognitive sciences. For instance, experimental psychology is naturally task oriented, because subjects are typically studied in the context of specific experimental tasks (see e.g., van Rooij, 2008). Notice that this understanding of cognitive tasks is very closely related to psychological practice. Therefore, cognitive tasks can be identified with functions from possible initial states of the world into possible final states of the world. Namely, the aim of a cognitive task is to transform the initial given state of the world into some desired final state. Taking a very abstract perspective we can say that a cognitive task is a computational task. One of the primary objectives of cognitive psychology is to explain human cognitive performance. We show that the computational distinction between quantifiers recognized by finite-automata (simple devices without internal memory) and push-down automata (finite-automata with data storage) is psychologically relevant.

In particular, we compare time needed for verifying different types of quantifier sentences. Therefore, we provide an argument in the recent debate on the role of computational complexity in the cognitive science (see e.g., van Rooij, 2008). Moreover, we give some direct empirical evidence linking the computational complexity predictions with cognitive reality.

We refer to a recent neuropsychological investigation of the same problem by McMillan, Clark, Moore, Devita, and Grossman (2005) and account for some troubles with the interpretation of its results (see Szymanik, 2007). We investigate the comprehension of simple quantifiers in natural language as described in a computational model posited by many linguists and logicians (see e.g., van Benthem, 1986).

Quantifiers reading comprehension pdf