The Knowledge-Representation Hypothesis

Last Update: 12 January 2005

Note: or material is highlighted

(From Smith, Brian Cantwell (1982), "Prologue to "Reflection and Semantics in a Procedural Language" (PhD Dissertation, Department of Electrical Engineering and Computer Science, MIT, Feb., 1982)," in Ronald J. Brachman & Hector J. Levesque (eds.) (1985), Readings in Knowledge Representation (Los Altos, CA: Morgan Kaufmann): 31-39.)

Version 1 (p. 33):

part 1:

"Any process capable of reasoning intelligently about the world must consist in part of a field of structures, of a roughly linguistic sort, which in some fashion represent whatever knowledge and beliefs the process may be said to possess."

"must": this is a necessary condition of intelligence.
"consist": the process consists of these representations
- i.e., they're really there
- contrast behaviorism, Dennett's "intentional stance"
the representations are "roughly linguistic": contrast connectionism
they represent "knowledge & beliefs": So, KR as: representation of knowledge; not: representation of world.

part 2:

"There is ... an internal process that ... "computes with" these representations."

part 3:

"This ... process ... react[s] only to the "form" or "shape" of these mental representations without regard to what they mean or represent."

i.e., the process is syntactic

Version 2 (p. 33):

"Any mechanically embodied intelligent process will be comprised of structural ingredients that a) we as external observers naturally take to represent a propositional account of the knowledge that the overall process exhibits and b) independent of such external semantical attribution, play a formal but causal and essential role in engendering the behavior that manifests that knowledge."

"mechanically embodied": including us? excluding angels?
"we as external observers":
- Dennett again;
- i.e., the representations aren't (necessarily) there?
- or: that the structures represent propositional information is an external (i.e., by us) semantic attribution
"propositional": i.e., linguistic
- but NB: not "object-oriented" (i.e., not a taxonomic hierarchy ontology)? (If so, then this rules out some standard KR languages!)
the structures play a causal role; so, they must be there--but that's a hypothesis!
- not sufficient for them merely to be there
- they have to be computationally realizable
- not epiphenomenal

William J. Rapaport (rapaport@cse.buffalo.edu)
file: 563S05/KR.hypoth-2005-01-12.html