I first wrote about the Philologos column of the Forward back in 2004, and once again they’ve come up with a nice bit of language history worth sharing in Recounting a Tale of Counting and Telling. It takes off from the quoted observation that Hebrew “sefer, ‘writing,’ ‘document,’ ‘book,’ and sofer, ‘scribe,’ ‘enumerator,’ ‘secretary,’ derive from one and the same verbal root s-f-r, meaning ‘to count, ‘to number,’ ‘to report,’ and ‘to recount,’” and goes on to “comment on the interesting fact that a verbal relationship between counting and narrating is not limited to Hebrew”:
Such a linkage exists in English, too — and not only in “count” and “recount,” two words mentioned by Labuschagne. We also find it in the verb “to tell,” which has the second, now archaic meaning of “to count,” as in a phrase like “to tell [the beads on] a rosary.”
Nor is English the only language that resembles Hebrew in this respect. German has zahlen, “to count,” and erzahlen, “to tell”; in Dutch this is tellen and vertellen; in Danish, taelle and fortaelle. All these languages belong to the Germanic branch of the Indo-European family — but a “count-tell” relationship is not restricted to it. In the Romance family of Indo-European, we encounter the same thing. In French, for example, compter is “to count,” and conter is “to tell” or “to relate.” In Italian it’s contare and reccontare. In Spanish a single verb, contar, means both “to count” and “to tell,” so that cuenta is a numerical reckoning or a bill, and cuento is a story.
[…] Let’s start with the Hebrew root s-f-r. “He counted” in Hebrew is “hu safar,” while “he told” is “hu sipper,” using the pi’el construction. Both are related to the Akkadian (old Babylonian) verb shaparu, whose original meaning was “to send,” but which in time came to mean “to send a letter,” and eventually, “to tell” or “to relate,” since this is what letters often do. The root has like meanings in other Semitic languages, but only in Hebrew did it take on the additional meaning of “to count,” which was clearly a later development.
In the Romance languages, on the other hand, the process was reversed. In classical Latin, computare — the source of our English “compute” — originally meant only “to count” or “to do sums.” Not until Late Latin, from which the various Romance languages evolved, did it take on the sense of “to relate.” Yet classical Latin had its own “count” — “tell” pair in enumerare, a verb that derived from numerus, “number” but also had the sense of “to narrate.”
And now for our third case: Old English tellan, the ancestor of our modern “to tell.” Its oldest meaning was “to count,” as it was in other Germanic languages, which later added the meaning of “to relate” with the help of prefixes like German –er and Dutch ver-. In addition, however, tellan in Old English also meant “to put [something] in order.” And that, of course, is the link between counting and telling. To count is to put numbers in their proper order, and to tell a story or relate an incident is to put events in their proper order, first things first and last things last. This is why the two things are associated in so many languages, including Hebrew.
Commenters point out that in Slavic, the root chit– means both ‘count’ (cf. Russian число < *chit-slo ‘number’) and ‘read’ (cf. Russian читать ‘to read’); it’s related to Sanskrit cit– ‘perceive, take note of’ (related to ketas ‘thought’). Thanks for the link, Paul!
German has zahlen, “to count,” and erzahlen, “to tell”
Oops. It’s zählen and erzählen. Zahlen is “to pay”, erzahlen gibbit nicht.
Finnish
lukea
(transitive) to read
(obsolete) to count, still used in some expressions like mukaan luettuna, mukaan lukien
Estonian
lugema (da-infinitive lugeda)
1. to read
2. to count, to matter
Sinu arvamus ei loe.
Your opinion doesn’t count.
Zahlen is “to pay”
Presumably related to English toll and talent, the currency.
Geeknote: Your blockquotes are no longer done with the <blocknote> tag, and they accordingly don’t work over RSS. If this is something that you, Moondog or John Cowan can adjust, the world would be a better place.
“Presumably related to English toll and talent”
Talent in fact from Gk τάλαντον, via Latin. OED gives this for ‘toll’:
As for 'compute', the put- stem originally denoted the pruning of trees, then assessing and reckoning, which I find quite an appealing sense-development.
We also find it in the verb “to tell,” which has the second, now archaic meaning of “to count,” …
The noun “teller” isn’t archaic though.
Geeknote:
<blockquote></blockquote>
(D’oh! Thanks, Stu.)
Surely the fixed phrase “all told” reflects the “count” meaning of “tell”. Though no doubt it is often understood as “when all has been said” rather than “when all has been added up”.
“Modern” French tonlieu? I had to look it up. The TLFI says this term belongs to le droit féodal, ‘feudal law’, which has not been current for a long, long time. The definition is very similar to that of l’octroi (masc), since both have to do with an internal customs system. One of my great-grandfathers spent his working life as an employee of the Paris octroi, which perceived duties on a number of commercial goods entering the city. This was still going on when he retired around 1930. What could have been a routine job was enlivened by the detection of pervasive fraud about the nature, quality and quantity of the goods subject to those duties.
Blockquote: I wonder why we need to use this cumbersome 10-letter word, instead of a single letter such as q (since b is already taken for ‘bold’).
the Paris octroi, which perceived duties on a number of commercial goods entering the city.
I love it ! Is “to perceive duties” still current legal terminology ? The OED marks “perceive” in this sense as obsolete. I wish “perceived duties” still meant money in my pocket, since I have a lot of them: duties that I imagine I have, but perhaps don’t really in the wider view of things, for instance to pay income tax.
marie-lucie: Blockquote: I wonder why we need to use this cumbersome 10-letter word, instead of a single letter such as q (since b is already taken for ‘bold’).
There is already a <q> or “q tag”. It is used to enclose a text in single or double quotations marks. <blockquote> does this, and also something additions. That additional thing depends (I think) on how it is “configured” by the “style sheet” info that comes with an html page.
“Something additional”
Your blockquotes are no longer done with the <blocknote> tag, and they accordingly don’t work over RSS.
Well, hell. I changed from blockquote to indenting because I thought the big quote marks looked stupid, but if the indent isn’t showing up in RSS maybe I’ll start using it again. Anybody else having/being bothered by/noticing this problem?
“Modern” French tonlieu? I had to look it up.
You have to understand that Conrad operates on a very different time scale than the rest of us. To him, Paracelsus is modern.
(I added blockquote and itals to Conrad’s comment for clarity; I hope he doesn’t mind.)
Another technical question: when I clicked on LH in the bookmarks toolbar of Firefox under the old system, it showed me the list of recent items so I could see if there was something new to read or not. Now it does not show a list a simply opens the blog. Is this inherent in the change, or could Songdog change it ?
I note that it is not now an RSS feed.
I note that it is not now an RSS feed.
It’s not?? Surely someone would have mentioned this to me before! At any rate, I’ll pass your comment on to Songdog and see what he says.
Juha: I am interested in the Finnish and Estonian words where the meaning count is associated with read. In the Tsimshianic languages of British Columbie (Canada) the same connection occurs.
The stem of the word for ‘count’ seems to be related to that for ‘hit, chop’, and I think that the semantic link refers originally to counting by hitting one’s fingers one by on on some support, such as one’s other hand, one’s knee, or whatever suitable object was at hand. (Early numbers were 1, 2, 3, 4, 5 and 10. Goods available for trade between tribes, such as skins or dried fish, were usually bundled together in multiples of ten). To this verbal stem is added a suffix meaning ‘by mouth’. I think that during the fur trade, accurate counting became especially important, intermediate numbers were coined and added to the language, and traders were heard counting with words rather than with their fingers: they were literally “finger-counting by mouth”. Not only that, but they were observed to say the counting words while looking at pieces of paper. With the coming of missionaries, who also read aloud off the page, the meaning of the verb was extended to any kind of reading. (The word for ‘write’ is derived from one meaning ‘to make designs’, a common semantic evolution).
The first sentence of Sefer Yetsirah says that God created the world bi-sfar v-sippur v-sefer, which as you can imagine has been translated in greatly varying ways.
Stu: perceived duties
I had been looking up tonlieu and octroi, just to make sure of the meanings, and since I was reading French sources I must have translated unthinkingly from the legal phrase percevoir des droits ‘to collect fees’. Would ‘collect’ be the right word? What do customs agents do?
Hat, your post still has an oops in it.
I don’t know what you mean by “still,” and I’d appreciate it if you’d just say what the oops is rather than making me hunt for it, which I do enough of in my day job.
As for ‘compute’, the put- stem originally denoted the pruning of trees, then assessing and reckoning, which I find quite an appealing sense-development
English “tally” is also supposed to be <= Lat. talea “twig, pruning”, as in, making notches on a twig or stick?
The stem of the word for ‘count’ seems to be related to that for ‘hit, chop’
probably paralleled by English “strike” in its meaning “count”, and especially by Russian “раз” “occurrence, count” related to words meaning “to strike, to hit” but also “to cut, to make notch”.
I contacted Songdog re RSS and he said “the new site still has an RSS feed, though the feed URL has changed. I suggest that he try deleting the bookmark, then bookmarking the site again. Maybe Firefox will detect the new RSS feed URL?”
Thanks, Dmitry. I knew about “tallying” numbers on a “tally stick” (no doubt also relatives of the “tell/toll” root), but had not thought of the possibility with the Tsimshianic word. But the custom of recording numbers with notches might have existed in North America. Does the “hit, chop/count” connexion occur among Siberian native people? Juha mentioned the connection in Finnish and Estonian. Intriguing.
Does Stu mean the omission of the umlaut in zahlen and erzahlen? If so, it is in the quote from Philologos.
I tend to imagine that the semantic commonality of number-“tell” and story-“tell” lies in reciting from memory. Being able to recite 1, 2, 3, … is more or less a prerequisite for working out how many things you have (and if you say the numbers aloud then this is the first thing an observer might notice about the process), and reciting a story that you have learned by heart is one way (a common way in ancient times) of telling a story.
Which is to suggest that the central idea is not merely “putting things in order” but rather “saying things in order by rote memory”.
Collect is correct.
Modern French begins around 1600, so even long-obsolete words can be part of it.
The only thing that comes to mind in Japanese is the (slightly) archaic “kataru,” to tell a story, whence “monogatari” = “tale” (lit. “the telling of the thing”) and “kazoeru/kazu,” to count/number. I can’t imagine what kind of sound change would result in t- and z- springing from the same origin, though. Maybe kant- becoming kant- + -u = kantu->kandu->kanzu->kazu (not unheard of) and kant- + -a = kanta->kata- ?? Not sure of that one. In that case one would expect kanta->kanda->kada.
The standard word, “hanashi,” comes from “hanasu”/”hanatsu” originally meaning to move away or to emit or to release, which it still does, as a homonym (話す = hanasu = tell/speak, 離す = separate/move apart, 放す = release/liberate). Nothing to do with counting.
Grumbly Hat: I don’t know what you mean by “still,” and I’d appreciate it if you’d just say what the oops is rather than making me hunt for it, which I do enough of in my day job.
I spelled it out in the very first comment on this thread.
And yes, the OED2 lists all told under sense 20 of tell, which is ‘to mention or name (the single members of a series or group) one by one, specifying them as one, two, three, etc.; hence, to ascertain from the number of the last how many there are in the whole series; to enumerate, reckon in; to reckon up, count, number’. Indeed, all told and tell beads/prayers are the only subsenses of that sense which are not marked obsolete, archaic, or dialectal, and I have my doubts about the second one: the entry is from 1911. There are some other cool phrases that have been lost, like tell N years ‘be N years old’, tell tears ‘weep’, and tell the clock ‘count the hours, waste time’.
Which is to suggest that the central idea is not merely “putting things in order” but rather “saying things in order by rote memory”.
But how could someone learn to say things “by rote memory” unless he already had a notion or practice of “rote” and “by heart”, and thus of repetition, and thus of order ? On the other hand, “saying things in order” is impossible to do and recognize without memory of some kind on the part of both speaker and hearer.
Memory, repetition and order follow each other as they develop, like the tigers in the tale (for butter or worse). This is a chicken and egg scenario. I don’t know why people fret about which came first, given that both are available at the local supermarket.
And butter too.
I spelled it out in the very first comment on this thread.
As Ø says, it’s in the quote from Philologos. I don’t change quotes. (Also, to me “an oops” means “an oops different from the one I already mentioned”; otherwise I’d expect “the/that oops.”)
“Account” is a good example of an English word (both as noun and verb) used both in a doing-arithmetic sense and a narrative-exposition sense.
Pokorny says the PIE root is del. From what I can make of the German, both zählen and zahlen came from this root, as do English talk and Greek δόλος list.
I don’t see it in the Pokorny entry, but AHD says Dutch taal language is from that root too, as is English tall.
FWIW, dil means language in Turkish. Anybody know its source?
It’s an RSS version 2 feed, which is licit even if unusual.
It’s from Old Turkic tıl, which is apparently from Proto-Turkic *til, *dɨl ‘tongue; language.’
FWIW, dil means language in Turkish. Anybody know its source?
It’s a native word for ‘language/tongue’, and its faint resemblance to IE *√dngh- is almost certainly a coincidence: one phoneme does not make a cognate. Sometimes this root is transcribed *dlngh- to account for Latin lingua, but since the older Latin form dingua is also recorded, it’s now generally thought that /d/ > /l/ under the influence of lingere ‘to lick’ < *√leigh- (presumably imitative).
“Paracelsus is modern”
I follow the good old fashioned tradition of referring to any authors who are post-classical as “recentiores”. Though in that case the “modern” was the OED’s word, not mine.
Another possible “tell”-“prune” link: Modern Hebrew siper, besides meaning “tell” as mentioned by Philologos, also means “give a haircut” (and from the same root, sapar “barber”, maspera “barbershop”). How old this sense is, or what its history is, I have no idea. (A consequence of this polysemy, combined with the indeterminacy of the script, is that the words for “literature” and “hair-cutting, the tonsorial art” are spelled the same, so if you’re signing up for a course in ספרות, it’s probably wise to call first.)
since the older Latin form dingua is also recorded, it’s now generally thought that /d/ > /l/ under the influence of lingere ‘to lick’ < *√leigh-
But there are other instances of so-called ‘Sabine l‘ in Latin, i.e. l where d is etymologically expected, and they can’t be accounted for this way: lacrima : Gk. dákru, oleō : odor, and a handful of others. As often in this kind of scenario, the fairy dust of dialect borrowing has been invoked, hence the term.
Apologies to Songdog for the misspelling his name (I’m totally on a roll today). Blockquote is totally the Right Thing; if the stupid big quote is the problem, it could be edited out of the stylesheet. (I am sufficiently familiar with css that I could undertake to do said editing and send you the result: it is in the very silly “lockquote::before” parts of the file accessible to us as http://languagehat.com/wp-content/themes/languagehat/style.css.)
I use http://languagehat.com/index.rdf for RSS; I think I didn’t change it with the switchover; it now redirects (transparently) to http://languagehat.com/feed/.
Sometimes this root is transcribed *dlngh- to account for Latin lingua, but since the older Latin form dingua
So English ‘language’ and English ‘tongue’ share the same PIE root. What goes around comes around.
Another possible “tell”-”prune” link: Modern Hebrew siper, besides meaning “tell” as mentioned by Philologos, also means “give a haircut” (and from the same root, sapar “barber”, maspera “barbershop”). How old this sense is, or what its history is, I have no idea.
The sense of hair-cutting would appear to be post-biblical. BDB dedicates several pages (start at 706 here) to safar, sefer, siper, etc. only in the senses of counting and recounting.
Hebrew Wiktionary gives examples of the haircutting sense. The earliest is from the Mishna.
JC: about “Modern French tonlieu”
Modern French begins around 1600, so even long-obsolete words can be part of it.
Tonlieu is listed in the TLFI, a dictionary which gives oldest sources and dates, like the OED. The entry starts with a paragraph indicating that the word belongs to le droit féodal ‘feudal law’, which has not been in use for centuries. In this para the word is included in a sentence from a comparatively recent book about history which mentions it as a type of tax collected by a nobleman at some point in the Middle Ages. The next para gives two quotes, both dating from the 12th century, with examples in Old French. This para is preceded by abbreviations including “Ac.”, before the number 1762. Perhaps this means that the word was considered by the Académie some time in 1762, using the citations later included in the TLFI. In any case, even though the word is mentioned in a 20C text, and perhaps in the archives of the Académie, it is only attested in its context of use in Old French, about 400 years before the acknowledged beginning of Modern French.
Stu, my point might be clearer if I say that I am distinguishing between (1) putting things in order and (2) saying things that one has memorized (in an order that one has memorized).
Of course I am not maintaining that one can say things in order without having an idea of order. I am merely speculating about the semantic range of a word.
Of course, as soon as I start using notions like “semantic range” or “word” with you there is a chance that the conversation will go off the rails. I remember that one occasion you told me that I was making words sound like parking spaces and meanings like cars — or was it the other way around?
@J. W. Brewer: Yes! I had the same thought. Specifically I was imagining that the expression “to give an account of oneself” is often right in the middle between storytelling and bookkeeping.
To “tell the clock” may be obsolete but to “tell time” certainly isn’t. Funny the OED doesn’t mention it- is it sense 22 “to reckon up or calculate the total amount or value of (money or other things); to count?”
Hebrew ספר spr ‘cut hair’, according to Klein, is an Aramaic loan; and compare Arabic شَفْرة šafra ‘razor, blade’. Jastrow’s Sefer Millim pulls together the semantics, explaining that both ultimately come from a root meaning ‘to cut’, and hence ‘mark’, ‘write’, ‘count’ and ‘tell’. This is also the root of Biblical Hebrew סְפָר ‘boundary’, and hence ‘district’. Seems reasonable to me, but I haven’t seen the nitty-gritty of the sound changes.
I used to see Moondog on the corner of 6th Ave. and 53rd.
http://en.wikipedia.org/wiki/Moondog
I’ve never been able to wrap my head around the Semitic sibilant correspondences, but according to this Wiki chart, Hebrew ס /s/ corresponds to Arabic س /s/ and to Akkadian /s/, so how can ספר be cognate with either شَفْرة šafra or Akkadian shaparu, anyway?
empty: Of course, as soon as I start using notions like “semantic range” or “word” with you there is a chance that the conversation will go off the rails.
Well, if it’s your train of thought then it’s your responsiblity to stay on track and ignore distractions. What I said here is in line with the discussion, on my view, though one perhaps couldn’t tell by counting the words. It’s actually rather profound, and summarizes in a jocular way ideas familiar to philosophers and cognitive scientists.
If you are more comfortable with prefound thoughts, that’s fine. I never insist.
I remember that one occasion you told me that I was making words sound like parking spaces and meanings like cars — or was it the other way around?
That’s pretty cool ! I don’t remember saying it, but I will.
On-topic: I’ve written a note to Philologos about the typos.
Off-topic (you have been warned): the Forward is an interesting newspaper, I find. Here’s the off-topic bit that might be topical for empty, another article in the Forward: Math and Anti-Semitism Went Hand-in-Hand at Harvard for Decades. It’s a review of a book about Birkhoff’s activities at Harvard.
TR, good catch. If /spr/ ‘cut hair’ is a borrowing from later Aramaic into Hebrew, you can posit Proto-Semitic *śpr ( is [ɬ]), which fits with the Arabic as well. But this doesn’t explain how the same root, with the meaning ‘count’ etc., would end up in Biblical Hebrew as ספר spr rather than שׂפר śpr. Although Hebrew ś and s merge later on, spr ‘count, recount’ appears as early as Genesis.
To add to what Juha posted, the Finnish verb “kertoa” means both “to tell” and “to multiply numbers.” A possible semantic loan from Indo-European?
Math and Anti-Semitism Went Hand-in-Hand at Harvard for Decades. It’s a review of a book about Birkhoff’s activities at Harvard.
We await you plug for Movies and Anti-Catholicism; Hand-in-Hand in Hollywood for Decades.
Marie-Lucie:
The question of whether tonlieu is a word of Modern French or not seems to be an offshoot of the great LH “What is an English word?” debate. Consider the English (indeed, native) word infangthief. It refers to the right of certain landowners to summarily execute thieves who were captured “hand-having or back-bearing” (that is, the stolen articles) on the landowner’s property. (If they had stolen things from his land but were captured elsewhere, the landowner would need the much less common right of outfangthief to summarily execute them.) This right ceased to exist in England around the beginning of the 14C, except as part of the local law of Halifax in Yorkshire, which guillotined such thieves right up to the mid-17C.
Now the word, unlike the practice, is by no means obsolete: it is, indeed, the only word that legal and social historians of the period use. If it is a very obscure word, that’s because not too many people study mediaeval legal history. The OED, which has not yet updated the word, has quotations for it right up to five years before the original OED1 publication of the word; Google Ngrams, which begins in 1800, shows a massive spike in 1868 (why, I don’t know) and lesser ones around 1925 and 1965. So, is infangthief a word of Modern English? I’d say so: it is part of the technical register of certain historians. You, perhaps, would disagree.
I forgot to mention a point which seems important: the spelling infangthief is a modernization, as the period spellings were infangthef and infangtheof. But this may just be another version of “double entente is a non-English expression that exists only in English”, a particular instance of the aforementioned great debate.
The question of whether tonlieu is a word of Modern French or not seems to be an offshoot of the great LH “What is an English word?” debate.
Well, a parallel, anyway. Looking at that thread, I still shake my head in wonderment at the number of people who seem to think, or unconsciously assume, that the English they know is all the English that’s needed, and any words they find strange or unnecessary are ipso facto not part of English. (And yes, infangthief is obviously part of English.)
LH, I did not say that tonlieu was not a French word, I just don’t think that it qualifies as a part of Modern French.
No, no, I wasn’t talking about you and tonlieu! I was just reminiscing about the earlier thread.
The only thing that comes to mind in Japanese is the (slightly) archaic “kataru,” to tell a story, whence “monogatari” = “tale” (lit. “the telling of the thing”) and “kazoeru/kazu,” to count/number.
Au contraire! There is a better example of the phenomenon: the verb that in contemporary Japanese means “read” originally meant “count”. MYS 4492:
都奇餘米婆 伊麻太冬奈里 之可須我尓 霞多奈婢久 波流多知奴等可
tukwi yomeba/ imada puyu nari/ sika su ga ni/ kasumi tanabiku/ paru tatinu to ka
“Counting the months/ It is still winter/ And yet/ The drifting mist:/ Can spring be here?”
This meaning is even in the dictionary the Jesuits compiled in the 1600s, and it turns up in Edo literature. I don’t recall seeing it used as a non-archaism in anything post-Meiji, though. (And I see from looking in the 日本国語大辞典 that as early as the 1200s it was being proscribed as an ill-bred (下賤) and uncouth (げす) way to say “kazoeru”.)
(Oh, and of course “yomu” originally meant, implicitly, “read aloud“; if you think of the verb as meaning not “absorb information from printed text” so much as “speak a provided sequence of items aloud one by one” the relationship to counting becomes clearer.)
Matt: if you think of the verb as meaning not “absorb information from printed text” so much as “speak a provided sequence of items aloud one by one” the relationship to counting becomes clearer
This is exactly what I think is the semantic link I was mentioning earlier (Jan 17 @ 9:18) as a comment to Juha’s contribution.
Thinking again about the Tsimshianic counting words, I think that I made the error of not recognizing that there were two separate problems: the names of the numbers (ome of which relate to a method of counting with hand and finger motions) and the verb meaning ‘count, read’. The meaning of the root word I interpreted as ‘hit’ refers to ‘hitting violently’ and is therefore closer to ‘chop’. The root therefore would not be a suitable one for finger counting but would be more appropriate to making notches in wood. The shape of the verb suggests that it was coined in the earliest recoverable period of the language, so it could not be due to Russian influence (at a time when there were Russian trading posts on the Coast, just North of the people I am talking about).