Explore the vast range of titles available.

A major goal of linguistics and cognitive science is to understand what class of learning systems can acquire natural language. Until recently, the computational requirements of language have been used to argue that learning is impossible without a highly constrained hypothesis space. Here, we describe a learning system that is maximally unconstrained, operating over the space of all computations, and is able to acquire many of the key structures present in natural language from positive evidence alone. We demonstrate this by providing the same learning model with data from 74 distinct formal languages which have been argued to capture key features of language, have been studied in experimental work, or come from an interesting complexity class. The model is able to successfully induce the latent system generating the observed strings from small amounts of evidence in almost all cases, including for regular (e.g., aⁿ, (ab)ⁿ, and {a, b}⁺), context-free (e.g., aⁿbⁿ, aⁿbn+m , and xxR ), and context-sensitive (e.g., aⁿbⁿcⁿ, aⁿbmcⁿdm , and xx) languages, as well as for many languages studied in learning experiments. These results show that relatively small amounts of positive evidence can support learning of rich classes of generative computations over structures. The model provides an idealized learning setup upon which additional cognitive constraints and biases can be formalized.

The first part of this article gives a brief overview of the four levels of the Chomsky hierarchy, with a special emphasis on context-free and regular languages. It then recapitulates the arguments why neither regular nor context-free grammar is sufficiently expressive to capture all phenomena in the natural language syntax. In the second part, two refinements of the Chomsky hierarchy are reviewed, which are both relevant to the extant research in cognitive science: the mildly context-sensitive languages (which are located between context-free and context-sensitive languages), and the sub-regular hierarchy (which distinguishes several levels of complexity within the class of regular languages).

Seki et al. (Theor. Comput. Sci. 88(2):191–229, 1991 ) showed that every m -multiple context-free language L is weakly 2 m -iterative in the sense that either L is finite or L contains a subset of the form , where w 1 ⋯ w 2 n ≠ ε . Whether every m -multiple context-free language L is 2 m -iterative, that is to say, whether all but finitely many elements z of L can be written as z = u 0 w 1 u 1 ⋯ w 2 m u 2 m with w 1 ⋯ w 2 m ≠ ε and , has been open. We show that there is a 3-multiple context-free language that is not k -iterative for any k .

Formal language theory (FLT), part of the broader mathematical theory of computation, provides a systematic terminology and set of conventions for describing rules and the structures they generate, along with a rich body of discoveries and theorems concerning generative rule systems. Despite its name, FLT is not limited to human language, but is equally applicable to computer programs, music, visual patterns, animal vocalizations, RNA structure and even dance. In the last decade, this theory has been profitably used to frame hypotheses and to design brain imaging and animal-learning experiments, mostly using the ‘artificial grammar-learning’ paradigm. We offer a brief, non-technical introduction to FLT and then a more detailed analysis of empirical research based on this theory. We suggest that progress has been hampered by a pervasive conflation of distinct issues, including hierarchy, dependency, complexity and recursion. We offer clarifications of several relevant hypotheses and the experimental designs necessary to test them. We finally review the recent brain imaging literature, using formal languages, identifying areas of convergence and outstanding debates. We conclude that FLT has much to offer scientists who are interested in rigorous empirical investigations of human cognition from a neuroscientific and comparative perspective.

We show how some problems in additive number theory can be attacked in a novel way, using techniques from the theory of finite automata. We start by recalling the relationship between first-order logic and finite automata, and use this relationship to solve several problems involving sums of numbers defined by their base-2 and Fibonacci representations. Next, we turn to harder results. Recently, Cilleruelo, Luca, & Baxter proved, for all bases b ≥ 5, that every natural number is the sum of at most 3 natural numbers whose base-b representation is a palindrome (Cilleruelo et al., Math. Comput. 87, 3023–3055, 2018). However, the cases b = 2, 3, 4 were left unresolved. We prove that every natural number is the sum of at most 4 natural numbers whose base-2 representation is a palindrome. Here the constant 4 is optimal. We obtain similar results for bases 3 and 4, thus completely resolving the problem of palindromes as an additive basis. We consider some other variations on this problem, and prove similar results. We argue that heavily case-based proofs are a good signal that a decision procedure may help to automate the proof.

We explore the generative capacity of morphological theories of reduplication. We computationally classify theories of reduplication using a hierarchy of string-to-string function classes. Reduplication as a process requires only the regular class of functions. We show that various morphological theories necessarily treat it as a more expressive polyregular function, while others maintain regularity. We discuss the significance of this formal result for reduplicative functions and recognition.

DNA is often described as the “language of life” because it encodes biological information using nucleotide sequences. Unlike the traditional view focused on codon-to-amino acid mapping in coding regions, the vast non-coding genome reveals complex organizational patterns resembling natural language. This paper outlines essential approaches in DNA linguistics, including formal language theory, RNA secondary structure modeling, statistical methods, and phylogenetic analysis. Additionally, recent research on Indo-European populations shows correlations between lexical and phonemic traits and asymmetrical patterns of genetic inheritance. Together, these perspectives deepen our understanding of genome regulation, evolution, and the striking parallels between genetic and linguistic systems.

The separability problem for word languages of a class$\\mathcal{C}$by languages of a class$\\mathcal{S}$asks, for two given languages$I$and$E$from$\\mathcal{C}$ , whether there exists a language$S$from$\\mathcal{S}$that includes$I$and excludes$E$ , that is,$I \\subseteq S$and$S\\cap E = \\emptyset$ . In this work, we assume some mild closure properties for$\\mathcal{C}$and study for which such classes separability by a piecewise testable language (PTL) is decidable. We characterize these classes in terms of decidability of (two variants of) an unboundedness problem. From this, we deduce that separability by PTL is decidable for a number of language classes, such as the context-free languages and languages of labeled vector addition systems. Furthermore, it follows that separability by PTL is decidable if and only if one can compute for any language of the class its downward closure wrt. the scattered substring ordering (i.e., if the set of scattered substrings of any language of the class is effectively regular). The obtained decidability results contrast some undecidability results. In fact, for all (non-regular) language classes that we present as examples with decidable separability, it is undecidable whether a given language is a PTL itself. Our characterization involves a result of independent interest, which states that for any kind of languages$I$and$E$ , non-separability by PTL is equivalent to the existence of common patterns in$I$and$E$ .

We consider nondeterministic higher-order recursion schemes as recognizers of languages of finite words or finite trees. We propose a type system that allows to solve the simultaneous-unboundedness problem (SUP) for schemes, which asks, given a set of letters A and a scheme G, whether it is the case that for every number n the scheme accepts a word (a tree) in which every letter from A appears at least n times. Using this type system we prove that SUP is (m-1)-EXPTIME-complete for word-recognizing schemes of order m, and m-EXPTIME-complete for tree-recognizing schemes of order m. Moreover, we establish the reflection property for SUP: out of an input scheme G one can create its enhanced version that recognizes the same language but is aware of the answer to SUP.

A function on an algebra is congruence preserving if, for any congruence, it maps pairs of congruent elements onto pairs of congruent elements. We show that on the algebra of binary trees whose leaves are labeled by letters of an alphabet containing at least three letters, a function is congruence preserving if and only if it is polynomial.