To Understand, or To Be Understood?

writing: to understand or to be

A common criticism of academic writing is that it is a chore to read and understand. The language seems to be purposefully perplexing, with abstract ideas mired deep within nested technical word chains, leaving you hanging with explanations that are terse and unforgiving.

Though, I think there's a good explanation for why people would want to write in such a style. It's got to do with the concept of abstraction, which basically allows us to think and reason at a higher level than the literal objects we want to reason about.

To simplify things, we should think of abstraction as a shift in our focus from objects to the ideas which they represent.

It really is that simple though, and we use it every day. For example, a bank statement is an abstraction of our combined transactions within a given timeframe, which allows us to reason at the bank account level. At this level, it makes sense to say that we are in an overdrawn account state, something which cannot be captured just by looking at a receipt for gas from the gas station (assume cash). Therein lies the power of abstraction.

But of course, we use different abstractions for different purposes. In the case above, we use a bank statement for the purpose of accounting. But we could just as well capture our transaction information with a different abstraction. Say for example, a spreadsheet of all your gas purchases, for the purpose of calculating your average fuel efficiency over the past year.

Now you're asking, "Who the hell does that?". And well, probably not many people would. But the point here is that some people would, and more importantly, that you could do things that way. In other words, abstractions can be specific to the purpose, and even specific to one's individual thought process. I hope that you can see how that is a potential recipe for complicating things really fast.

So abstraction is insanely useful, but strictly in the empirical sense. This is good, and some even argue that there can never be a high enough level of abstraction (or at least, within the bounds of our current knowledge of computational complexity).

But in most of our everyday dealings, we're not aiming for empiricism; we're aiming for comprehension. And unfortunately, in the attempt to get other people to comprehend what we say, we cannot use abstractions which only make sense to those who understand them.

In the case of our bank statement, we use our understanding of 2-dimensional rows and columns, along with our understanding of addition and subtraction, along with our understanding of the layout of a bank document to come to the conclusion that the bank manager is revelling in sadistic glee as he watches our overdraft fees mount. Trying to explain that to a child (without the necessary background knowledge) is going to be pretty hard.

But of course, this is precisely the pitfall which academics fall into all the time, mainly because once you've understood an abstraction, the simple act of re-writing the succinct string of words used to capture that abstraction makes one shudder with an inexplicable sense of joy. There is then a compelling urge to force this joy upon others; You become convinced that other people NEED to know what you just captured in a succinct abstraction.

I think it's important that we be aware of this pitfall, and avoid it whenever we catch ourselves swaying too far to the "academic" style.

However, the wise would now see the opportunity presented by our argument: By writing in an "academic" style, you generate new abstractions. By then reverting to a more "common" style, you break down those abstractions into their components. This cycle forces one to approach a particular abstraction from varying angles each time we (re)commence it.

Needless to say, in writing this post, I was required to bounce back between abstract and non-abstract concepts, as well as "academic" and "common" linguistic representation. In doing so, I too approached this topic in ways that I didn't, nay, couldn't, think of before. ie: I've used abstraction to capture abstraction! =P Paul Graham is right in saying that "writing doesn't just communicate ideas, it generates them."

In other words, get to writing! But don't forget that while it's great (and necessary!) to write to understand, we should only present to others what we wrote to be understood.


comments powered by Disqus