And the traditional way to count newspaper columns (for payment of reporters on casual rates), was by the column inch. By the 1980’s some newspapers and magazines had switched to character counts for casual rates, but I don’t remember if it was /5 or /6
This is my thinking as well. Clear logic. +1
FWIW, I will mention that my wife was a professional translator, French to English. She charged by the word of the source text. A phrase such as qu’est-ce counted as three words.
Helping to demonstrate that unless there’s a particular explicit purpose for counting some words, there can’t be clear answers to questions on how it ought to be done.
Back in the day when telegrams were a thing (I sent a few in the 1980s, usually for birthdays, one when the party did not answer their phone), words of more than 10 characters counted as one word per 10 characters. People economised by dropping pronouns, prepositions and articles, and merging separable verbs (ANKOMME MONTAG ELFUHRDREI = I will arive on Monday at 11:03 am).
Nowadays it’s flat rates of 12.90€ for up to 160 characters, 18.35€ for up to 480 characters, for telegrams within Germany, but of course they are not transmitted over teletype anymore.
Whoosh.
Readability measures such as Flesch count contractions as one word. (Cite is a 1.2MB PDF, see p. 8, 26)

FWIW, I will mention that my wife was a professional translator, French to English. She charged by the word of the source text. A phrase such as qu’est-ce counted as three words.
And shows how wanting to maximize income drove the definitional process.

Whoosh.
Got me. In my defense pulykamell had a typing error that through* me off. That’s a type of error I do all the time so it stands out for me.
* I leave that as an exercise for the student.
Thanks for the replies everyone. Actually I did not have a particular purpose of counting in mind, it was more of whether there is a universally applied social convention of considering a contraction as one or two words conceptually. It seems to me that the predominant view is that, unless a particular counting purpose dictates otherwise, it would be considered as one.

Got me. In my defense pulykamell had a typing error that through* me off. That’s a type of error I do all the time so it stands out for me.
Are we speaking of human-fingered typos, rather than erroneous auto-completions? If so, are we speaking of the aphasia I’ve mentioned before, with vowels–>values as my go-to example, though I’ve made other, even more bizarre, substitutions?
If so, I admit to noticing these in other Dopers’ posts, though only since I first mentioned my peculiar aphasia. Sorry if it was contagious and I was the Typhoid Mary. :eek:

Because-someone-will-always-come-along-and-claim-hyphenated-words-count-as-a-single-word …
Well, time period enters into that also.
I’ve worked with Distributed Proofreading for Project Gutenberg, dealing with books written a century or more ago. It’s common to see terms like to-day or to-morrow as hyphenated items. In even older works, they are two completely separate words to morrow.
I’ve even seen times given as 10 of the clock or 10 o’ the clock.
Not sure how those would be counted as words. I don’t very often see much use for counting words.

I don’t very often see much use for counting words.
I have six words for you: See the quotes in the OP

The standard was five characters plus a space, which is why you’ll see variously 5 or 6 used. If you didn’t know what the number meant you weren’t a writer and so could be ignored.
Thanks for the info. I must of misunderstood about the space due to my computing background where a character is a character.