Proofreading using a text-to-speech engine

I do my proofreading using a text-to-speech engine. Sometimes, I use Dragon NaturallySpeaking to create a document or a blog post. Other times I type it out. But either way, I like to have the computer read it back to me (NaturallySpeaking makes this very convenient). Listening to someone (or a computer) read my words out loud helps me catch things that I would otherwise miss. Sentences that sound great in my head sound awkward on being read out loud. Hearing them lets me catch awkward and incoherent sentences. More importantly, I hear typographical errors and incorrect words even when I don’t see them.

This was one of the most frustrating things in writing papers in graduate school: I would write something, submit it for review, and discover that there were little errors that really hurt my credibility. For instance, I would repeat repeat a word. I did proofread. I read through my documents carefully. I even read them out loud to myself. And yet, I would read the sentences as I had intended to write them and not as I had actually written them.

I had an experience with a student that showed just how insidious this could be. There was a question on my test that the student asked about. Part of the lab course was on safety and most of the safety questions on the test were very easy.

55.  What can reduce the probability of accidents?

A. use the largest quantity of material possible to accomplish the goal of the experiment

B. when possible, substitute a more hazardous chemical for a less hazardous one

C. skip planning the experiment and “wing it”

D. anticipate the possible consequences of the work you do in the laboratory

The student asked, “aren’t B and D both good answers?”

I asked him to read answer B again. He read it out loud to me as written. And yet he processed it mentally to mean the opposite of the written statement. Nobody would intentionally use a more dangerous substance instead of a safer alternative (that’s why B is a wrong answer). It’s so obviously not something that anyone would want to do, that the mind balks at reading it that way. In fact, the student read the words in the correct order but processed the meaning in the opposite order.

My point is not the student was at all foolish. In fact, he was very sensible: his mind refused to accept that anyone would be so stupid as to do this stupid thing. Here’s the problem: if one is proofreading, or taking a test, and looking for wrong things, it is not an advantage to read things this way. The task is to look for the incoherent statements. If we subconsciously substitute in coherent statements (instead of reading literally) then we will not be able to find the incoherent statements.

This is a great reason to use a text-to-speech engine. It seems like when I hear the words read to me, I am not reading them in my own voice in my head. As a consequence, I seem to be a great deal more critical. And in proofreading, that is key to success.

%d bloggers like this: