Columnist: Computer voice-recognition technology currently cannot distinguish between homophones such as "their" and ...

Irina on July 13, 2019

@Shiyi The assumption we are looking for is that the voice recognition program must be able to distinguish between homophones to produce accurate translation, and to distinguish between homophones, it needs to recognize and utilize grammatical and semantic relations among words.
In other words:

recognize grammatical and semantic relations -> distinguish between homophones -> accurate translation.

Answer choice D is incorrect because it flips steps 1 &2 saying that in order to recognize grammatical and semantic relations among words, the technology must distinguish between words that are homophones, i.e. distinguish between homophones -> recognize grammatical and semantic relations.

Let me know if this helps.

Devon on February 20 at 05:27PM

Hi Irina - I'm not clear on the above diagramming. Wouldn't "in order to recognize grammatical and semantic relations among words, the tech must distinguish between homophones" be diagrammed as "recognize grammatical and semantic relations -> distinguish between homophones," since distinguishing between homophones is necessary? Similarly, wouldn't all the diagrams above be flipped? Like, in order to be a doctor, you must go to med school?

Emil on February 22 at 12:10AM

Hi, I don't think I'm understanding your question, she diagramed that statement as you did. D is incorrect because it flips them.

Devon on February 22 at 09:33PM

lol - very sorry Irina and company - too many hours staring at the screen