Summary Of Steven Pinker's Talking Heads

In Steven Pinker’s Language Instinct chapter 7 “Talking Heads”, he explains the difference between human language acquisition and the ability for computers to perform the same. In the past, many people feared that computers would end up conquering humans due to the development of “artificial intelligence” in the 1950s. Research has debunked this myth with the simple fact that for computers “the hard problems are easy and the easy problems are hard” (191). For example, a simple question like “Do zebras wear underwear?” can confuse the computer. This suggests that the computers could eventually successfully replace the harder jobs like engineering rather than the more simple jobs like a receptionist. Pinker goes into more depth on why this is …show more content…
People usually use the depth-first strategy. The depth-first parser works by “picking an analysis that seems to be working and pursing it as long as possible; if they come across words that cannot be fitted into the tree, they backtrack and start over with a different tree” (212). Therefore, the human parser uses a top-down strategy where they assume the intentions of the speaker at the beginning and everything else in the sentence is used to understand that first thought. Computers use a breadth-first search. A breadth-first search completely eliminates any unlikely phrases or ambiguous words in order to understand the sentence. People usually don’t do this because “many sensible ambiguities are simply never recognized” (210). Humans using the depth-first search typically causes problems when understanding garden path sentences. In garden path sentences, their first words tend to lead the reader to an incorrect analysis. With garden path sentences, it is clear to see that humans do not build all possible trees like computers do, but assume there is one true tree structure at first and learn they are wrong when they reach a word that doesn’t fit that