Publications

In this section you can find an annotated list of publications that will guide you to the papers that are most interesting for you depending on how familiar you are with FCG and what your interests are. The papers are therefore ordered according to three themes:

1. Introductory papers: If you are new to the formalism, then it is a good idea to start with these papers. You may also want to read the theoretical background on this website and the notes with each publication to get a general overview of FCG.
2. Implementation papers: These are the publications that give formal descriptions of the machinery of Fluid Construction Grammar. These are therefore more technical than the others and are mostly useful for researchers who want to get a full grasp of FCG or even expand it to their own needs.
3. Papers on experiments and methodology: These papers report on the experiments that have been conducted with FCG and on the methodological issues that are relevant for exploring new research topics. These are highly recommended for everyone who wants to discover the possibilities of FCG and wants to set up new experiments. Some familiarity with FCG is useful, but these papers can be read without having gone through the more technical publications.

Introductory Papers

Steels, L. and De Beule, J. (2006) A (very) Brief Introduction to Fluid Construction Grammar. [PDF]

A very dense informal overview of the current state of FCG (as to may 2006) that was prepared for the Third International Workshop on Scalable Natural Language Understanding (ScaNaLU 2006) June 8, 2006, following HLT/NAACL, New York City.

Steels, L. (2005) The Role of Construction Grammar in Fluid Language Grounding. Submitted. [PDF]

This paper examines in how far Construction Grammar is a useful foundation by which artificial agents can self-organize communication systems that are grounded in the real world through a sensori-motor embodiment and use grammar to express certain aspects of meaning. It also gives an introduction to Fluid Construction Grammar. Note: In this paper, templates are called rules.

Steels, L. (2004) Fluid Construction Grammars. A Brief Tutorial. [PDF]

This document is a bit outdated with regard to some implementation details, but still is a good first read to grasp the working of Fluid Construction Grammar (an updated manual is in preparation).

Implementation Papers

Steels, L. and De Beule, J. (2006) Unify and Merge in Fluid Construction Grammar. [PDF]

In: Lyon, C., Nehaniv, L. and A. Cangelosi (eds.), Emergence and Evolution of Linguistic Communication, Lecture Notes in Computer Science, Springer-Verlag: Berlin.

The unify and merge operators form the core of the FCG formalism. This paper explains how unify and merge use the templates of a speaker's language inventory in a bi-directional way to parse or produce constructs.

De Beule, J. and Steels, L. (2005) Hierarchy in Fluid Construction Grammar. [PDF]

In: Furbach, Ulrich (ed.), Proceedings of the 28th Annual German Conference on AI, KI 2005, Lecture Notes in Artificial Intelligence (vol. 3698), pages 1-15, Berlin Heidelberg.

Early implementations of FCG dealt with single layered structures without hierarchy. It is obvious however that natural languages make heavy use of hierarchy in different levels of organization. Introducing hierarchy without the loss of bi-directionality has been one of the hardest challenges of the past two years, but an elegant solution has been found in the J-operator. The key idea to handle hierarchy is to construct a new unit as a side effect of the unify and merge processes. This paper goes into detail on the working of the J-operator and shows how it can handle hierarchy both in the syntactic and semantic domain.

Steels, L., De Beule, J. and Neubauer, N. (2005) Linking in Fluid Construction Grammars. [PDF]

In: Proceedings of BNAIC, Transactions of the Belgian Royal Society of Arts and Sciences, pages 11-18, Brussels.

The linking problem addresses one of the key issues in any language processing system: how to establish an adequate syntax/ semantics interface. In other words, how can constructions be linked to each other to express more complex utterances? In FCG, the linking of constructions is licensed by construction templates. This paper is a first exploration of this process.

Papers on Experiments and Methodology

Steels, L. (2006) How to do Experiments in Artificial Language Evolution and Why. [PDF]

In: Cangelosi, A., Smith, A. and K. Smith (eds.), Proceedings of the 6th International Conference on the Evolution of Language, World Scientific Publishing: London.

This is a paper that focuses on methodology issues. It examines a few examples, evaluation criteria and what can be expected from these kind of experiments.

Steels, L. (2005) The Emergence and Evolution of Linguistic Structure: From Lexical to Grammatical Communication Systems. [PDF]

Connection Science, 17(3-4), September-December, p. 213-230.

This paper deals with the modeling of language from a cognitivist angle. It does not discuss new experiments, but provides a research roadmap for everyone who is interested in setting up experiments in this field.

Steels, L. (2004) Constructivist Development of Grounded Construction Grammars. [PDF]

In: Daelemans, Walter (ed.), Proceedings of the Annual Meeting of Association for Computational Linguistics, 2004.

This paper gives some examples of experiments as well as an introduction to the formalism as it was in the beginning of 2004. Many of the issues raised in this paper have already been dealt with in more detail and published elsewhere (e.g. hierarchy) or are in preparation (e.g. an experiment on marking perspective reversal in language).