Jordi Fortuny

Diversity, competition, extinction: the ecophysics of language change

As indicated early by Charles Darwin, languages behave and change very much like living species. They display high diversity, differentiate in space and time, emerge and disappear. A large body of literature has explored the role of information exchanges and communicative constraints in groups of agents under selective scenarios. These models have been very helpful in providing a rationale on how complex forms of communication emerge under evolutionary pressures. However, other patterns of large-scale organization can be described using mathematical methods ignoring communicative traits. These approaches consider shorter time scales and have been developed by exploiting both theoretical ecology and statistical physics methods. The models are reviewed here and include extinction, invasion, origination, spatial organization, coexistence and diversity as key concepts and are very simple in their defining rules. Such simplicity is used in order to catch the most fundamental laws of organization and those universal ingredients responsible for qualitative traits. The similarities between observed and predicted patterns indicate that an ecological theory of language is emerging, supporting (on a quantitative basis) its ecological nature, although key differences are also present. Here, we critically review some recent advances and outline their implications and limitations as well as highlight problems for future research.

Emergence of Zipf’s law in the evolution of communication

Zipfs law seems to be ubiquitous in human languages and appears to be a universal property of complex communicating systems. Following the early proposal made by Zipf concerning the presence of a tension between the efforts of speaker and hearer in a communication system, we introduce evolution by means of a variational approach to the problem based on Kullbacks Minimum Discrimination of Information Principle. Therefore, using a formalism fully embedded in the framework of information theory, we demonstrate that Zipfs law is the only expected outcome of an evolving communicative system under a rigorous definition of the communicative tension described by Zipf.

Towards a mathematical theory of meaningful communication

Meaning has been left outside most theoretical approaches to information in biology. Functional responses based on an appropriate interpretation of signals have been replaced by a probabilistic description of correlations between emitted and received symbols. This assumption leads to potential paradoxes, such as the presence of a maximum information associated to a channel that creates completely wrong interpretations of the signals. Game-theoretic models of language evolution and other studies considering embodied communicating agents show that the correct (meaningful) match resulting from agent-agent exchanges is always achieved and natural systems obviously solve the problem correctly. Inspired by the concept of duality of the communicative sign stated by the swiss linguist Ferdinand de Saussure, here we present a complete description of the minimal system necessary to measure the amount of information that is consistently decoded. Several consequences of our developments are investigated, such as the uselessness of a certain amount of information properly transmitted for communication among autonomous agents.

On the Origin of Ambiguity in Efficient Communication

This article studies the emergence of ambiguity in communication through the concept of logical irreversibility and within the framework of Shannon’s information theory. This leads us to a precise and general expression of the intuition behind Zipf’s vocabulary balance in terms of a symmetry equation between the complexities of the coding and the decoding processes that imposes an unavoidable amount of logical uncertainty in natural communication. Accordingly, the emergence of irreversible computations is required if the complexities of the coding and the decoding processes are balanced in a symmetric scenario, which means that the emergence of ambiguous codes is a necessary condition for natural communication to succeed.

Introduction. On the locus of ambiguity and the design of language

The general objective of this thematic issue is to bring together researchers working within different fields, such as experimental linguistics, artificial intelligence, categorial grammar, theoretical linguistics, sign languages and complex systems, in order to provide a comprehensive and up-to-date theoretical study on the presence of ambiguity in natural languages. The idea of editing an interdisciplinary volume that focused on the importance of ambiguity for the study of language originated from our previous works Fortuny and Corominas-Murtra (2013), Corominas-Murtra et al. (2011), CorominasMurtra et al. (2014) – the two last articles are developed in collaboration with Ricard Solé. These three intimately related articles investigate how conditions of communicative efficiency can be used to understand certain properties of linguistic productions. In Fortuny and Corominas-Murtra (2013) we show that if we take into consideration the complexities of both coding and decoding processes and we assume that communicative agents tend to minimize these two processes, then a certain amount of ambiguity must emerge. We express this tendency in terms of a symmetry equation between the coding and the decoding complexities that allows us to quantify the minimal amount of ambiguity that must appear. As a result, the presence of ambiguity derives from certain efficient considerations about natural communication. In Corominas-Murtra et al. (2011) we capitalize on the symmetry equation in order to show how it may yield Zipf’s law. The crucial point is that in order to obtain Zipf’s law as the unique outcome of the symmetry equation we need to consider how a code evolves through time. More precisely, we need to ensure that code evolution is path-dependent. This requirement is naturally expressed in terms of the Minimum Discrimination of Information Principle: the most likely code at step n + 1 is the one minimizing the distance with respect to the code at step n. Therefore, Zipf’s law derives from two economy

The Witness Set Constraint

This article is initially concerned with a famous constraint on the class of possible determiners in natural languages: the so-called Conservativity Constraint. We shall briefly illustrate the force of this constraint and informally sketch Keenan and Stavi (1986)’s view according to which the Conservativity Constraint derives from the boolean structure of natural language semantics. We shall proceed to discuss certain well-known linguistic categories that have been argued to be interpreted by non-conservative functions: only and the relative proportional determiners many and few. We shall take the challenge posed by the existence of these categories in order to propose an alternative to the Conservativity Constraint. This alternative will be dubbed the Witness Set Constraint, which is inspired in Barwise and Cooper (1981)’s considerations on the semantic processing of generalized quantifiers. We shall defend that the proposed constraint does not suffer from the empirical shortcomings that have been attributed to the Conservativity Constraint, and indeed, we shall argue in detail that it correctly predicts (a) the existence of conservative determiners, (b) the non-existence of certain non-conservative determiners, such as inner negations, cardinal comparison determiners and the converses of non-trivial proportional determiners, and most importantly, (c) the existence of the non-conservative functions denoted by only and the relative proportional determiners many and few. This line of reasoning suggests that the class of functions from properties to sets of properties denoted in natural languages typically by determiners is constrained by a principle that simplifies the semantic processing of generalized quantifiers.