Introduction: How to Build a Talking Robot - Computational Linguistics and Talking Robots - page 7

1.3.2 C OMPARISON OF LANGUAGE , CONTENT , AND PATTERN PROPLETS

German proplet French proplet content proplet pattern proplet

⎡
⎣

⎤
⎦

⎡
⎣

⎤
⎦

⎡
⎣

⎤
⎦

⎡
⎣

⎤
⎦

sur: Hund
noun: dog
cat: m-g
sem: sg
fnc: bark
mod: old
prn: 23

sur: chien
noun: dog
cat: sn m
sem: sg
fnc: bark
mod: old
prn: 24

sur:
noun: dog
cat: sn
sem: sg
fnc: bark
mod: old
prn: 25

sur:
noun: α
cat: sn
sem: sg
fnc:
mod:
prn: K

Proplets for a word in different languages differ mainly in their sur (face)
value, here Hund vs. chien . Language and content proplets differ in the pres-
ence or absence of a sur value. Pattern and content proplets differ in the pres-
ence or absence of values defined as variables, here and K. In addition there

is the distinction between lexical proplets and connected proplets. The latter
represent content resulting from hear mode derivations (e.g., 3.3.1), nonlan-
guage recognition (Chap. 8), and inferencing (Chaps. 5, 6, 10, 11). 13

α

1.4 Time-Linearity

The development of DBS may be traced back to a Montague-style formal
“fragment” of natural language, published as SCG'84. 14 The goal was to
realize the First Principle of Pragmatics (PoP-1) 15 in a system of Truth-
Conditional Semantics. This gave rise to the following question:
Is the agent part of the model, or is the model part of the agent?
For analyzing climate change, pollution, the stock market, etc., the first option
is appropriate. For reconstructing cognition, the second option must be taken.

12 A precursor of feature structures in linguistics is a list of binary values without attributes, called a “fea-

, used by Chomsky and Halle (1968) for purposes of morphophonology.
The introduction of feature structures with attribute-value pairs is credited to Minsky (1975). Called
“frames” by Minsky, their origin may be traced back to the legendary Dartmouth workshop of 1956,
attended also by Newell, Simon, McCarthy, Rochester, and Shannon.
13 In the JSLIM software (Weber et al. 2010), proplets always have all their attributes in the appropriate
order (cf. NLC'06, Sect. 13.1). In this text, however, proplets are handled in a more liberal fashion:
they are often shown with only a subset of their attribute-value pairs. The purpose is a more succinct
presentation and drawing attention to properties relevant to the issue at hand.
14 Thanks to N. Belnap and R. Thomason at the Philosophy Department of the University of Pittsburgh
(1978-1979), and to J. Moravscik and G. Kreisel at the Philosophy Department of Stanford University
(1979-1980, 1983-1984) for revealing the secrets of a Tarskian semantics. Later, at Carnegie Mellon
University (1986-1989), understanding this fundamental topic was helped further by Dana Scott. Cf.
FoCL'99, Chaps. 19-21. The 1978-1979 stay in Pittsburgh and the 1979-1980 stay at Stanford were
supported by a two-year DFG research grant.
15 PoP-1: The speaker's utterance

ture bundle”, e.g., +vocalic
+high

meaning 2 is the use of the sign's literal meaning 1 relative to an