INVERSE BOOK REVIEWS. — Hooray. ... [ Word Count: 900 ~ 4 PAGES | Revised: 2018.6.4 ]

tibra (58)in #science • 6 years ago (edited)

Book review incoming.

[WIN86] Terry WINOGRAD, Fernando FLORES, Understanding computers and cognition, Norwood: Ablex, 1986.

I'll be writing some inverse paper and book reviews.

What's that? An inverse review ... is what?

Here's the gist.

— 〈 1 〉—

Some deep books require unpacking. Everybody recognizes that.

Others, equally deep, however, require more than unpacking. They require deeper and deeper packing.

Huuuhhh?

Paradoxically one thing becomes less opaque when being unpacked, but another becomes less opaque when being packed.

After all the goal is comprehension [CHA03]. Which is also the reason why packing or unpacking differently affect understanding.

Some descriptions requires unpacking: in order to later repackage and compress in another way.

The compression appropriate to the reader is what makes comprehension [CHA03]. And that is understanding.

It's why we seek simple laws that explain complicated or complex appearances.

Some descriptions require summary in ordinary language. Deformalization.

Others descriptions however require formalization: mathematical reviewing.

— 〈 2 〉—

Language often hides the abstraction used [MIL67]; X is not always X.

Just to demonstrate how vague our ordinary language is in an obvious example. "Metal" cannot always replace "Metal". A door knob cannot replace a coat hanger, but both are "Metal" and our long verbal arguments are strings of this is that, which in turn is such and such ... and that in turn is something else. Misleading.

When it concerns something everybody already understands ... errors and misunderstanding are few ... And if sometimes they are not few ... then at least they're obvious to just about everybody.

Not so when the subject matter is abstract. When most people do not already understand it.

So abstraction is as often a source of error and control as enlightenment in ordinary language.

We use mathematical language to reveal all encapsulation and other kinds of abstraction.

An inverse book review will begin with a mathematical restatement of the points or arguments in the reviewed work.

Often different and even conflicting arguments turn out, in another language, to be really the same, with some minimal and explicit abstraction.

Making the abstraction explicit then gives a single strand to focus and talk about in ordinary language, and makes the subject more, rather than less accessible.

— 〈 3 〉—

Sometimes a technical monograph or paper requires unpacking in ordinary language to make it more clear to more people. Other times however it requires compressing into mathematical language to make it more clear; and this is one of those times.
WIN86. This is a deep book.

I'll try to unpack it by giving a shorter mathematical summary. Unpack it by first packing it. (Lovely.)

Then after that:

[ECC80] John ECCLES, The human psyche, Berlin: Springer, 1980.

I keep planning on discussing [BOD06], but I find it better to cover some of the primary sources first.

Separate the large complicated and complex problems into simple, succinct problems that the reader can consume in arbitrary order. Consumed together, however, the large problem reappears, but is more clear.

The same principle as that it's easier and more likely to hit a target consecutively from various positions one or two feet away, in whatever order, a million times consecutively, than hit it twice from a hundred feet away.

What time is needed to comprehend a thousand simple problems that can be combined into the same large problem by their interactions is often less than to comprehend the really equivalent large problem.

But that, too, is in the future.

Preview: Pattern Recognition:

People are good at this kind of thing in effect ...