Making the task feasible: the problem of machine translation has not been fully solved, but we can achieve some limited subgoals:

· Rough translation (adequate for a general understanding of what a text is about): cross-language information retrieval

· Translation that will be improved by a (possibly monolingual) human editor (post-editing, "Computer-aided human translation")

· Small sublanguage domains (e.g. weather reports)

The transfer model: Starting from a structural analysis, use rules about differences between languages to translate directly from one surface structure to another: syntactic transformations (adjusting word order) and lexical transfer (selecting equivalents).

The interlingua model: Interpret into a semantic representation, then generate out from it.
Direct translation: a collection of ad-hoc algorithms that progressively change one sentence into another based on low-level information.

Statistical techniques: find correspondences between sentences that maximize fluency (probability of a sequence in the target language) and faithfulness (lexical matching based on aligned corpora).

Often you can figure out what kind of algorithm is being used by examining the errors.

Alta Vista French-English (a post to an e-mail list)

Original:

MERCI D'AFFICHER ET DE DIFFUSER / PLEASE POST

Si votre messagerie n'a pas un éditeur de traitement de texte WORD le même
programme avec une meilleure mise en forme et quelques couleurs se trouve
dans le fichier attaché (format Word 2000 sous Windows 98) – Si vous
souhaitez un envoi en fichier rtf ou sous une aure forme, n'hésitez pas à le
demander à crl1@libertysurf.fr <mailto:crl1@libertysurf.fr> / If
WORD is not the default editor of your mail box you will find the same
program attached ( WORD 2000 – WINDOWS 98) – If you need it as an rtf file
please ask for it at crl1@libertysurf.fr <mailto:crl1@libertysurf.fr>

ACQUISITION ET CONSTRUCTION DU SENS DANS UNE PERSPECTIVE INTERLANGUE

Colloque international sur les acquisitions tardives en langue maternelle

Amphithéâtre Louis Liard - Paris-Sorbonne - 20 &21 décembre 2001

Organisé par la Cellule de Recherche en Linguistique (CRL) avec le soutien
du Laboratoire d'Étude sur l'Acquisition et la Pathologie du Langage chez
l'Enfant (LEAPLE, CNRS-UMR 8606 - Université Paris V), et la Cellule de
Recherche Fondamentale en Linguistique Française et Comparée (CRFLFC,
composante de l'Équipe d'accueil 2283 de l'Université de Franche-Comté).

Translation:

THANK YOU TO POST AND DIFFUSE/PLEASE POST If your transport does not have an editor of text processing WORD the same program with the best not formatted and some colors finds in the attached file (Word format 2000 pennies Windows 98). If you wish a sending in file rtf or under a aure forms, do not hesitate to ask it crl1@libertysurf.fr < mailto:crl1@libertysurf.fr >/If WORD is not the default editor of your mall box you will find the same program attached (WORD 2000. WINDOWS 98). Yew you need it have year rtf spins please ask for it At crl1@libertysurf.fr < mailto:crl1@libertysurf.fr > ACQUISITION AND CONSTRUCTION OF SENS FROM THE POINT OF VIEW INTERLINGUA International conference on late acquisitions in mother tongue Amphitheatre Louis Black poplar - Paris-Sorbonne - 20 &21 December 2001 Organized by the Cell of Research in Linguistics (CRL) with the support of the Laboratory of Study on the Acquisition and the Pathology of the Language at Child (LEAPLE, CNRS-umr 8606 - University Paris V), and the Cell of Fundamental Research in French and Compared Linguistics (CRFLFC, component of the Team of reception 2283 of the University of Franche-Comté).

Alta Vista English-Spanish

Original:

Does anyone know of any case studies in first language acquisition where a
child reverses first and second person, referring to himself in second
person and to his interlocutor in first person? (Such a practice dispenses
with the deictic element and allows second person to be an absolute
designation for this particular child.)

--Aya Katz

Translation:

¿Cualquier persona sabe de estudios de caso en la primera adquisición de la lengua adónde un niño invierte primero y segunda persona, refiriéndose en la segunda persona y a su interlocutor en la primera persona? (tal práctica dispensa con el elemento deictic y permite que la segunda persona sea una designación absoluta para este niño particular.) -- Aya Katz