Name a technical term with Latin or Greek

March 20, 2007 3:01 PM

Latin and/or Greek experts, I need help naming a technical term.

I am writing an academic paper concerned with the timing of digital logic. In a digital system, data signals are usually synchronized to a central clock which oscillates at a fixed frequency. A higher frequency means a faster "speed". In larger systems there can be several different clocks - for example in a computer your processor, memory, and peripherals are all clocked at different speeds.

Group of clocks can be categorized according to their timing relationships to one another. A well-known paper has done so and invented the following now well-known terms:

- mesochronous: exactly the same frequency, but possibly out of phase.

- plesiochronous: nominally the same frequency, but with a slight mismatch (eg. 1Ghz and 1.00000001 Ghz)

- heterochronous: nominally different frequencies.

There is a fourth term we would like to add. It is a subset of heterochronous where the frequencies are different, but are exact rational multiples of one another. For example, if there are three clocks with frequencies of 100Mhz, 200Mhz, and 500Mhz, then the second is a factor of 2 faster than the first, and the third is a factor of 5 faster than the first, and the third is a factor 5/2 faster than the second.

This term is normally called "rationally related clocks" or "rational clocks", where rational means 'a ratio of two integers'^. But no term is in wide use yet and it would be nice if we can find something that fits into the aforementioned blank-chronous form. I expect this will be tough because we did sort of hunt around for a good term and came up emtpy. But it's worth a shot. Any ideas?

posted by PercussivePaul to writing & language (30 comments total)

On the greek side, hemi-, iso-, and para- could all be shoehorned, though none is great; I find parachronous the most appealing of the three.

Latin: metrochronous, implying something of the discrete multiples?
posted by cortex at 3:12 PM on March 20, 2007

although why the hell I'm trying to suggest 'metro' is Latin and not Greek I do not know.
posted by cortex at 3:14 PM on March 20, 2007

In music, a polyrhythm is similar to what you describe. Perhaps Polychronous would be the right word?
posted by deadmessenger at 3:15 PM on March 20, 2007

mensurochronous, with the intention of evoking commensurability.
posted by Wolfdog at 3:19 PM on March 20, 2007

Allochronous, riffing on allomorphic.
posted by adamrice at 3:19 PM on March 20, 2007

Or, I suppose, even comensurochronous; that's becoming a bit unwieldy, though.
posted by Wolfdog at 3:20 PM on March 20, 2007

Something based off of "harmonic", as in wave harmonics?
posted by Khalad at 3:33 PM on March 20, 2007

I like parachronous myself ("para-" meaning alongside or related); I also thought of zygochronous for similar reasons ("zygo-" meaning yoked together).

The suffix "-metric" works but doesn't match the other models; the prefix "metro-" usually means urban (metrosexual) or uterine (metritis).
posted by rob511 at 3:33 PM on March 20, 2007

I am pleasantly surprised to see so many answers already but nothing quite fits (told you it was tough!). Comensurochronous is the best so far and seems to be as precise as we are likely to get, but I fear it is indeed too unwieldy. I will run it by my co-authors, though.

I think we actually came up with 'para' and 'poly' and decided they weren't sufficiently descriptive, precise, or different from the existing terms. We also considered standbys like 'iso' and 'semi' and 'pseudo' and decided that they were too vague, not to mention overused in the literature as it is.
posted by PercussivePaul at 3:49 PM on March 20, 2007

Ratiochronous seems to capture the idea of rationally related clocks rather well. And ratio is Latin.
posted by harmfulray at 4:39 PM on March 20, 2007

Khalad: "Something based off of "harmonic", as in wave harmonics?"

I think this might be the most descriptive. Harmochronus? Harmonochronus? They certainly don't slide off the tongue, do they?
posted by JMOZ at 4:57 PM on March 20, 2007

The problem with ratiochronous is precisely that ratio- is Latin, while -chronous is Greek. Anyone know the Greek for ratio?
posted by nomis at 4:58 PM on March 20, 2007

(Interestingly, the mathematical term irrational does come from Greek: αρρητος, "unspeakable" or "inexpressible", and despite the tempting appearance of "ratio" in the English form, it does not seem to be related to that concept.)
posted by Wolfdog at 5:09 PM on March 20, 2007

To stay within the Greek combining family, how about quasisesquichronous? prefix "quasi-" (resembling), prefix "sesqui-" (one-and-a-half)
posted by rob511 at 5:11 PM on March 20, 2007

mensurochronous

im sorry i didnt get to this post ealier, but this really made me snortle.
posted by phaedon at 6:35 PM on March 20, 2007

let me also add in response to:

- mesochronous: exactly the same frequency, but possibly out of phase.


well, *exactly* the same frequency should be called idiochrono(u)s. but i also found this definition online:

mesochronous: The relationship between two signals such that their corresponding significant instants occur at the same average rate. (my emphasis)

this seems right.

plesiochronous

im not that familiar with that many words that use plesio from the modern greek verb plesiaso this way. its a strange way of putting it. an ancient greek root would be peri- as in, peripatetic or periphery. perichronous.

to actually answer your question. here's one solution:

to use a root such as hyper- or super- to signify signals that are hyperchronous by a factor of three. (the same with subchronous. please keep in mind i havent checked for overlapping definitions in the audio world.)

i was going to recommend something like pentachronous in an effort to outline the relationship of a factor of 5 (borrowing from the pentatonic concept in music theory) but i got myself all confused and abandoned the idea.

To stay within the Greek combining family, how about quasisesquichronous? prefix "quasi-" (resembling), prefix "sesqui-" (one-and-a-half)

quasi- and sesqui- aren't in the greek family!
posted by phaedon at 6:50 PM on March 20, 2007

I showed this thread to my supervisor (one of my co-authors). He was intrigued by comensurochronous. Neither of us knew the mathematical definition of comensurate, which seems to apply quite well in this case. Ratiochronous he kind of shrugged off; it's a decent word in my opinion, but maybe it sounds too artificial, probably because ratio is normally an English word (and maybe the Latin-Greek mixture has something to do with it). Those are the best so far. There is still a third author I have yet to show it to.

Quasisesquichronous is fun to say but too wordy for anyone to take seriously. The harmonic idea is interesting but I don't know how to turn it into a snappy word.

Still open to any more ideas. Thanks for the help so far!

On preview, in response to phaedon:
our definitions of mesochronous are equivalent in a practical sense. and while you may be right about plesiochronous, this is the term the original author came up with years ago and we are stuck with it now. And your suggestions aren't quite what I'm looking for; when describing specific examples one would use English words as I did. We only need the Greek word to describe the category in general.

On further preview:
I googled 'ratiochronous' and found that one paper in 2005 actually used this term in precisely this way, though it looks like they just invented it themselves. This paper is not well-known enough that I or my co-authors have read it. Still, it counts as a precedent, and that may be enough.
posted by PercussivePaul at 7:11 PM on March 20, 2007

how about polydiplochornous, polypentachronous? oh. i see. your're looking for a more general term. hm...

how about metachronous - as, "occuring at different frequencies"?

or better yet - anachronous - from the greek analogos - analogochronic? - same as the Latin for proportion - as in, "referring to things that share a similar relation"..

or: polychronous. bleh, weak.
posted by phaedon at 7:33 PM on March 20, 2007

pollaplasichronous or pollaplusichronous. Or polyplusichronous.

Too long? Pollaplasio (πολλαπλάσιο) is excatly this: the multiplicative constant.
posted by carmina at 9:43 PM on March 20, 2007

Too long? Pollaplasio (πολλαπλάσιο) is excatly this: the multiplicative constant.

What I really meant is:

Too long? Pollaplasio (πολλαπλάσιο) is exactly: multiple.
posted by carmina at 9:52 PM on March 20, 2007

hot dog, i second that!
posted by phaedon at 9:56 PM on March 20, 2007

Wow, you guys are smart. Ratiochronous probably wins out due to precedent, but I will present the highlights to my other co-author and see what he says. Check back later for results. Thanks.
posted by PercussivePaul at 10:48 PM on March 20, 2007

Please do not use "comensurochronous." Words that mix Latin and Greek look horrible to anyone who knows any classical languages, and that one in particular is long and unwieldy as well; at least "ratiochronous" (though also an ugly mix) is shorter and easier to say. The Greek word logos can be used for 'ratio, proportion' (e.g., 'in proportion to' is ana logon); how about "logochronous"? Or phaedon's "anachronous," although that would have unfortunate interference from "anachronism."
posted by languagehat at 5:09 AM on March 21, 2007

"Harmonically related" would have meaning to people in every field of engineering, it is absolutely precise (it would seem), and it is just as easy to write/say as these compound words you are looking at..
posted by Chuckles at 4:21 PM on March 21, 2007

Well, no, harmonically related is not absolutely precise, because the harmonic series is a specific set of fractions and in this case it could be any two integers. 'rationally related' is I think as precise as you can get.

In any case it looks like ratiochronous wins due to precedent, as I suspected.

I don't think we would have used comensurochronous, it was too unwieldy. and really ratiochronous is the only term that unambiguously describes the situation; the only interpretation that springs to mind is some sort of ratio. Everything else could be interpreted in a number of ways.

thanks for the help everyone.
posted by PercussivePaul at 6:56 PM on March 21, 2007

Whether they are harmonically related or not would depend on the whether the base frequency is present, I guess. Would a family of clock rates like 200MHz, 500MHz, 700MHz, fall into the category you wish to define?

because the harmonic series is a specific set of fractions and in this case it could be any two integers.

They way I read this, it is exactly opposite to the actual meaning of harmonic :)
posted by Chuckles at 8:00 PM on March 21, 2007

yes they would. in practice there has to be a stable clock signal to generate the family, but you can divide as well as multiply, so the base could be, say, 7 Ghz (unrealistic in this case, but you get the idea). In general harmonics simply don't come to mind when I think about this concept.

What I meant by the latter sentence is that the harmonic series is 1/2, 1/3, 1/4, 1/5 ... and I want to allow 2/3, 7/6, x/y in general. The phrasing was sloppy though.
posted by PercussivePaul at 10:18 PM on March 21, 2007

Well, once you get away from the basic physics of a vibrating string, and start talking signal analysis, harmonics are integer multiples. From a signal analysis perspective, the fractions are just a consequence. That is kind of semantic though, I have to admit :P

On searching, harmonically related is used a fair amount, but the only definition I could find related to TV broadcast. I did see it used for clock signals a couple of times as well (one was talking about clock prorogation, saying that even though clocks were "harmonically related" when generated, they can easily be out of phase somewhere down the line)..

but you can divide as well as multiply,

This kind of makes my whole point moot, although I have a suspicion that it might be possible to massage something out of it anyway..

Now I'm thinking ratiometric, but those linguists up thread won't like it :)
posted by Chuckles at 8:55 AM on March 22, 2007

that would be "clock propagation"...
posted by Chuckles at 8:56 AM on March 22, 2007

Interesting. I can see how harmonics would come to mind in something like TV broadcasting because you think of signals as delta functions in the frequency domain. I think the reason it doesn't occur to me in this sense is that I think of my clocks as square waves in the digital domain, and their frequencies are chosen only by the maximum frequency of the logic they control. Perhaps 'harmonically related' would be more intuitive than 'rationally related' to an analog designer. However, none of the papers I have read describing this particular concept use 'harmonically related', as far as I can recall; most use 'rationally related' or some variant.
posted by PercussivePaul at 2:48 PM on March 22, 2007