Blog Response #4



Understanding, Understanding…Are we smarter than a computer?

Processing Pinker Chapter 7

Pinker begins chapter 7 with an analysis of artificial intelligence and how it evolved into recreational tools that conveniently support everyday tasks. While reading the chapter I reflected on my experiences with computers and examine how I too use technology daily. For example, I refer to “Siri” for answers about the time, or simple questions such as “hey Siri how many ounces are in a cup”? While at work, I use online reservation software to manage family care reservations at my center. Even during a trip to the grocery store, I see how stores have replaced cashiers with advanced self-checkout devices that allow you to scan and pay for your own groceries. What’s more, some stores have computerized shopping carts that scan your items and that allow you to pay directly from your cart! With the new advancements in technology, we are seeing more instances where practical tasks completed by people are replaced with machines. Is technology taking over the world? 

Pinker delves into this idea and examines the purpose of technology. According to Pinker (2000), “the main lesson of thirty-five years of AI research is that the hard problems are easy, and the easy problems are hard” (pg.191). From what I gather from the reading “hard” problems are the tasks that we encounter daily. For example, locating the time, questioning how many ounces or in a cup, or in my case reserving care for the students at my center. It doesn’t go to say that humans are incapable of completing these hard tasks, it's that technology systems help support our ability to solve these problems. However, when it comes to an “easy” task such as comprehending sentence structure, technology appears inferior, and human intelligence appears to have the upper hand (well more or less). 

Our ability to understand language “works when the incoming conversation or text is structured in certain ways” (Pinker 2000, pg. 194). Unconsciously, we find the subject, object, and verbs and integrate grammar rules to comprehend the structure of the sentence. Next, we integrate our memory and decision-making to decide what rules should we implement to complete the sentence. Our processing ability to comprehend sentences is powerful, but not perfect and has limitations. One limitation that hinders our mental processing is if the sentence structure is incomplete. When sentences contradict our universal grammar, our mental process “can bog down, backtrack and misunderstand” (Pinker 2000, pg.194).

Likewise to Pinker, I’m curious as to why this response is so. Pinker states that the “only way to prevent onion sentences would be to claim that the mental grammar defines two different kinds of a noun phrase, a kind that can be modified and a kind that can go inside a modifier” (Pinker 2000, pg. 204). If in fact, language acquisition is innate, we deduce universal grammatical rules, and we have a limited parser memory load, then positing different kinds of noun phrases would complicate our mental grammar, thus making it harder to comprehend. In Pinker’s words, it “would give the child an exponentially larger number of rules to record when learning the language” (Pinker 2000, pg. 204). But what if in fact, our mental capacity is not specific to innate language rules alone? What if in fact the external rules and the language we hear around us also formalize our grammatical categories? Could this concept explain why children can learn a language and understand sentence structure through social interaction, thus, debunking Chomsky’s and Pinker’s assertion that language acquisition is innate? QTNA đŸ¤”đŸ¤”

References: 

Pinker, S. (2000). The language instinct: How the mind creates language. HarperPerennial. 



Comments