Distilling Text into Circuits

  • 2023-01-25 13:56:34
  • Vincent Wang-Mascianica, Jonathon Liu, Bob Coecke
  • 10


This paper concerns the structure of meanings within natural language.Earlier, a framework named DisCoCirc was sketched that (1) is compositional anddistributional (a.k.a. vectorial); (2) applies to general text; (3) captureslinguistic `connections' between meanings (cf. grammar) (4) updates wordmeanings as text progresses; (5) structures sentence types; (6) accommodatesambiguity. Here, we realise DisCoCirc for a substantial fragment of English. When passing to DisCoCirc's text circuits, some `grammatical bureaucracy' iseliminated, that is, DisCoCirc displays a significant degree of (7) inter- andintra-language independence. That is, e.g., independence from word-orderconventions that differ across languages, and independence from choices likemany short sentences vs. few long sentences. This inter-language independencemeans our text circuits should carry over to other languages, unlike thelanguage-specific typings of categorial grammars. Hence, text circuits are alean structure for the `actual substance of text', that is, the inner-workingsof meanings within text across several layers of expressiveness (cf. words,sentences, text), and may capture that what is truly universal beneath grammar.The elimination of grammatical bureaucracy also explains why DisCoCirc: (8)applies beyond language, e.g. to spatial, visual and other cognitive modes.While humans could not verbally communicate in terms of text circuits, machinescan. We first define a `hybrid grammar' for a fragment of English, i.e. apurpose-built, minimal grammatical formalism needed to obtain text circuits. Wethen detail a translation process such that all text generated by this grammaryields a text circuit. Conversely, for any text circuit obtained by freelycomposing the generators, there exists a text (with hybrid grammar) that givesrise to it. Hence: (9) text circuits are generative for text.


Quick Read (beta)

loading the full paper ...