PEP 312 - Making lambdas implicit worries me, surely it's just the name 'lambda' that is bad...

[Alex Martelli]

Stephen Horne wrote:
   ...
> What Jack was asserting was not just that lambda is a long keyword
> that is (almost) always in the middle of a line. His intention was to
> assert that this property is responsible for making the word 'lambda'
> "jar the eyes".

Ah, thanks for the clarification; I had indeed missed this intention.


> To me, it seems absurd that the ugliness of a word could be a function
> of whether it happens to be a keyword or some other kind of word. The
> distinction between keywords and identifiers has no relevance to how
> jarring something is when you read it.

Probably not; the keyword/identifier distinction is very important
in some contexts but presumably not in that of "jarringhood" (:-).


> BTW - on the 'never'/'rarely' issue - there is a term "relative
> absolute" which with a simple play on words expresses the fact that it
> is perfectly valid to use 'absolute' words as extreme relatives.

I'm not familiar with the specifical oxymoron you quote, but in more
classic terms it does seem a form of hyperbole, or even catachresis.
Like many other rhetorical figures, it can usefully color one's speech,
but particularly in technical discourse it's best used cautiously, as
it can engender misunderstandings.

In this specific case, since there ARE some keywords that can never
start a (logical) line ('and', &c), AND some that can but rarely do
('lambda'), conflating the two with a hyperbolic "never" can in
fact leave some doubt in the reader -- particularly when unsure of
the point of the whole discourse -- as to whether the writer is
aware of the distinction, and whether the distinction being thus
brushed away is or isn't relevant to the issue anyway.  I'm not sure
the confusion is worth the slight "coloring" (see, litotes [which is
here shading into aporia] can also offer one's discourse some color,
while being less prone to such misunderstandings as the more popular
hyperbole can often engender...).


> Superlatives are, quite simply, not always *strictly* superlative.
> It's all part of the subjectivity of natural language, and one of the
> reasons context can be so important - not to mention, where relevant,
> redundancy (in the error detection and correction sense).
> 
> Knowledge of these kinds of things is only really useful if you want
> to out-pedant a language pedant, of course ;-)

Knowledge of rhetorics has its uses, and providing hi-falutin' terms 
that are not widely known is one such use, particularly in pedanter-
than-thou duels...


Alex