gis

The Hebrew word for brother-in-law is גיס - gis, and sister-in-law is גיסה - gisa. Klein takes the conservative view, and writes:

Aramaic-Syriac גיסא, aphetic for אגיסא, which is of uncertain origin

He was probably following Ben-Yehuda, who also wrote that the origin of the word is unclear, but then added that

there are those who say that it comes from gis גיס in Aramaic, which means "side"

Gis as side is familiar from the Aramaic phrases chad gisa חד גיסא - "the one side" and idach gisa אידך גיסא - "the other side". Avshalom Kor (Yofi Shel Ivrit, page 84) agrees with this explanation, and writes that gis is a "relative on the side".

None of the modern dictionaries connect gis to gayis גייס - "troop", which is the source of giyus גיוס - "draft". For example, Klein only points out that gayis is related to the Arabic word jaysh - "army".

However, Jastrow has a more extensive etymology for gis that connects it to gayis as well. He says that both are from the root גוס, which means "to come in contact, touch, be connected, meet". The Aramaic verb גוס does mean "meet" in Pesachim 110b and Gittin 65b - I couldn't find anyone who disagreed with that. He then goes on to say that gisa גיסא meaning "neighborhood, side" derives from that root (I guess sides are connected to one another), and then connects brother-in-law to side, as we've seen before.

He connects gayis as well, by saying that it means "troop", but "especially ravaging troop, invaders, robbers". The connection here seems to be that an invader comes in contact with the area he invades.

But there is one thing I don't understand in Jastrow's theory. He also writes that the root גוס is the source of another meaning of גיס - "intimate, familiar". He brings examples from Ketubot 85b and Kiddushin 81a, where the phrase גיס ביה - "familiar with him" is used. It seems clear to me that this is just the Aramaic form of the Hebrew -gas ba גס בה. However, Jastrow says that the Hebrew gas comes from an unrelated root, meaning "to be bold", so gas ba means "he may be become bold towards her" (Ketubot 12a).

This etymology of gas isn't too far from Klein, who says that it originally meant "coarse, bulky" and then meant "vulgar, impolite, discourteous". But what caused Jastrow to miss the obvious connection between the Aramaic and the Hebrew? Was it his need to tie in gisa and gayis, and explain the meaning of gis as "meet"? Does anyone have a more charitable explanation?

neched and achyan

While some of the kinship terms we've looked at so far have changed their meanings over time, we haven't seen a basic term that needed to be "invented". However, in 1943, the Jewish National Fund (KKL) asked the Vaad HaLashon for a Hebrew term that to us might seem obvious - nephew.

In this letter, they write (my translation):

In the listings in the memorial books of the KKL, it is customary to note the relationship between the dedicators and the dedicatees, as in: So-and-so is recorded by his parents, brother, etc ... It is not clear to us if the word nin נין means the son of the grandson or granddaughter, which is the popular usage, but is opposed to the opinion of the commentators.

Also, as to the new words dodan דודן and achyan אחין - it is questionable whether to accept them, and if we do, what is their feminine form.

The question was put to a committee headed by Tur-Sinai. In response, they wrote a dictionary of family terms, which can be found on this site (unfortunately, I can't link directly to it - to access it, click on the link הצגת מלון on the right, and then in the drop down menu choose משפחה).

Most of the terms in the dictionary just have a simple translation - English to Hebrew / Hebrew to English. But a few of them have notes - with the longest one being for their translation of "nephew". They translate nephew as nechdan נכדן (and niece as nechadnit נכדנית) and write:

The committee suggests the word nechdan נכדן (not achyan אחין), on the basis of the fact that in our Medieval literature the word neched נכד had that meaning (nephew), and it is still used in the linguistic traditions of some ethnic groups. It is also worth pointing out that the Latin word nepos, from which are derived the words "nephew" and Neffe (German), is used for both meanings: the son of the son (or son of the daughter) and the son of the brother (or son of the sister). By adding a nun at the end (of neched), as was done with dodan דודן (cousin) from dod דוד (uncle), we achieve the necessary distinction needed today.

We'll discuss dod and dodan at a later time, but first a couple of points about their note. The word neched appears three times in the Tanach (Bereshit 21:23, Yishayahu 14:22, Iyov 18:19) - each time paired together in a phrase together with nin נין. We've already noted that while in Modern Hebrew nin means "great-grandson", in Biblical Hebrew it appears to mean "son". According to most commentators, neched in Biblical Hebrew does mean grandson. It doesn't appear much in Talmudic Hebrew (Jastrow offers first "offspring", second "grandson" - and only brings one example.) But the Medieval use of neched as nephew that Tur-Sinai mentions can be found in the Rashbam on Bava Batra 108a.

And to elaborate on the various meanings of the English word "nephew", here's the Online Etymology Dictionary entry:

c.1297, from O.Fr. neveu (O.N.Fr. nevu) "grandson, descendant," from L. nepotem (nom. nepos) "sister's son, grandson, descendant," in post-Augustan L., "nephew," from PIE *nepot- "grandchild," and in a general sense, "male descendant other than son" (cf. Skt. napat "grandson, descendant," O.Pers. napat- "grandson," O.Lith. nepuotis "grandson," O.E. nefa, Ger. Neffe "nephew," O.Ir. nia, gen. niath "son of a sister").

"Niece" derives from nephew:

1297, from O.Fr. nièce (12c.), earlier niepce, from L. neptia, from neptis "granddaughter," in L.L. "niece," fem. of nepos "grandson, nephew". Replaced O.E. nift, from P.Gmc. *neftiz, from the same PIE root. Until c.1600, it also could mean "a granddaughter" or any remote female descendant.

And the word nepotism is also connected:

"favoritism shown to relatives, esp. in appointment to high office," 1662, from Fr. népotisme, from It. nepotismo, from nepote "nephew," from L. nepotem (nom. nepos) "grandson, nephew". Originally, privileges granted to a pope's "nephew" which was a euphemism for his natural son.

According to Almagor-Ramon (Rega Shel Ivrit #19), the double meaning of the Latin term nepos influenced the Hebrew usage of neched in the Middle Ages.

However, the official suggestion of nechdan ran into trouble early on. In 1948 Avinery wrote (published in Yad Halashon, page 389) that we shouldn't be surprised that there was no word for nephew in Hebrew. Ancient people weren't as careful about family titles as we are today, and even terms like av אב, ach אח and dod דוד didn't always mean father, brother and uncle (respectively).

He continues to say that he agrees with the committee's rejection of achyan for nephew, but nechdan has not caught on either - either in speech or in literature. His suggestion? To use the word nin for nephew. On the one hand, he prefers the Biblical shilesh שילש - "one of the third generation" - for great-grandson. On the other hand, he brings a few examples from the 1930s where nin or nina נינה was used for nephew or niece. For example, in 1931, HaAretz referred to Lord Balfour's niece, Blanche Dugdale, as his nina.

But this suggestion also didn't stick. The common term used today for nephew is achyan, and achyanit אחיינית is used for niece. It's not clear to me when (or why) achyan gained its popularity. I asked a neighbor of mine, who's in his sixties, if he grew up using the word achyan. He said no, when he was growing up, he'd say ben-ach בן-אח, and that achyan has only been used in the past 30 years or so.

Despite its new usage as nephew, the word itself goes back to Biblical times - it was the name of a member of the tribe of Menashe, listed in Divrei HaYamim I 7:19. As with most Biblical names, there is no clear etymology. Some sources suggest it meant "brotherly or fraternal"; others offer "younger brother". Yehuda Kil in Daat Mikra mentions that "some say" that it means nephew. I'm very curious who suggested that theory, and whether they were influenced by the modern meaning of the word...

teomim

We've discussed ach and achot - brother and sister. Now lets look at a word for a subset of siblings - twins. The Hebrew word for twins, teomim תאומים - appears four times in the Tanach: twice in Bereshit (25:24, 38:27) and twice in Shir HaShirim (4:5, 7:4). The singular form - תאום teom - never appears. According to Ben-Yehuda, it never appears in the singular in Talmudic Hebrew either - but that's much harder for me to confirm, since I don't have a complete concordance of Talmudic Hebrew as I do for Biblical Hebrew. Jastrow does mention the singular feminine form - teuma תאומה - but not the singular masculine teom. Ben-Yehuda also notes that it isn't clear from the Biblical usage whether the word only meant two children born at the same time, or if it would be used for triplets or quadruplets as well.


The parallel verbal root תאם finds its way into a number of verbs:

• תאם - to be duplicate, to be similar, to resemble, to correspond
  
• תיאם - to coordinate, so tium תיאום is coordination
  
• התאים - to suit, fit, match. According to some, the Biblical use (Shir HaShirim 4:2, 6:6) meant "to give birth to twins". Jastrow writes that in Talmudic Hebrew "to be twin-like, joined, adjoining." Klein adds that in Medieval Hebrew, this form of the verb was causative, and meant "he fitted, suited, conformed, adapted". In Modern Hebrew it took on the passive form, and began to mean "was fit, was suited, was adapted". The expression matim li מתאים לי means "it suits me" or, better, "it works for me".

One English word is related to the Hebrew word teom: the name Thomas. The Online Etymology Dictionary shows how the name came from the cognate Syriac תאמא:

from Gk. Thomas, of Aramaic origin and said to mean "a twin" (John's gospel refers to Thomas as ho legomenos didymos "called the twin;" cf. Syriac toma "twin," Arabic tau'am "twin"). Before the Conquest, found only as the name of a priest. After 1066, one of the most common given names.

Horowitz (p 284) adds that:

Tom like Jack is used to indicate the male of the species. Thus we have "tom Turkey" or "tom cat", the male and tougher variety of those interesting animals. A "tomboy" is a girl who acts like a boy.

(It should be noted that there are other explanations for "tomboy" - there are those that connect it to the word "tumble", because the girl dances - tumbles - around like a boy.)

Interestingly, a popular name for boys in Israel is Tom תום (pronounced "tome"). I'm guessing that Israelis like it because it has only one syllable (very common in secular Israeli first names) but also because it sounds more like an English name. I wonder how many of them know that this word too has Hebrew origins...

ach and achot

Continuing in this series, let's look at the Hebrew words for brother and sister - ach אח and achot אחות. (The generic word for "sibling" is achai אחאי). Ach is certainly related to the verb אחה - meaning "to join, to stitch". Despite the fact that the verb only appears in post-Biblical Hebrew, Klein writes that there is debate amongst scholars whether the verb derived from the noun, or the noun derived from the verb. Steinberg expands the root meaning "to join" to other words - ach אח - "fireplace", where people gather around, and achu אחו "reeds" (later "meadow"), which were used to make ropes for binding. However, modern scholarship has determined that both of these words were borrowed from Egyptian.

In addition to "brother", ach can also mean "kinsman" or "friend", and in Medieval Hebrew - "friar". In addition to "kinswoman", achot can also mean "sweetheart, companion, mate". Parallel to ach meaning friar - achot took on the meaning "nun" in the Middle Ages. In Modern Hebrew it also means "nurse". Ben-Yehuda doesn't mention that in his dictionary; but this 1956 article mentions that:

Under the influence of the German and British use of the word "sister" for "nurse," the Hebrew is now using the word achot (sister) also to mean "nurse."

(Interestingly, the Wikipedia entry for nurse writes that "the professional courtesy title 'sister' has fallen into disuse and disapproval, even though it was formerly used by both male and female registered general nurses.")

Achot is an unusual word. Despite it being a singular noun, it ends in -ot, which is usually the plural suffix of feminine nouns. We found a similar phenomenon in another kinship term: chamot חמות - "mother-in-law". How did this develop?

According to this book, the words ach and cham חם, as well as av אב, originally had three letters, not two. The last letter was a vav (in some forms changed to a yud), and can still be seen in the smichut (construct) form - avi- אבי, achi- אחי, chami- חמי (e.g. avicha אביך - "your father", and not אבך, achiv אחיו - "his brother", not אחו) as well as the familiar Abu prefix meaning father in Arabic. According to the author, the original form of ach was אחו (similar to the Akkadian ahu), which became achot אחות in the feminine. The last letter in the masculine form dropped off, but we still have the strange ending in the feminine.

Ben Yehuda explains it slightly differently - he says the original form of achot was achat אחאת, which is אחא with the tav added on as the feminine suffix (I assume he's connecting the word to the Aramaic אחא - acha "brother"). He writes that already in ancient times the patach ("a") changed to a cholam ("o").

Whether or not the original ach had three letters, there are a number of other words in Biblical Hebrew ending in -ot that are actually single, not plural. Here are some that I've found (some commentaries explain them otherwise):

• מכות - Micha 1:9
• חכמות - Mishlei 9:1
  
• בהמות - Iyov 40:15
• עדות - Tehilim 132:12

Each of these may have a different explanation, but it does show us that our initial assumption that -ot at the end of a word means plural is not always correct.

One common mistake by new Hebrew speakers (I remember making it myself) is to say achot when you mean "sisters". The correct term in Hebrew is achayot אחיות, but those who make the mistake are actually close to ancient Hebrew. David Talshir of Ben Gurion University wrote an article called "The Forms ‘ahot and ‘edot in Ancient Hebrew" where he discusses the plural of achot. The absolute form of the plural never appears in the Tanach. Achayot does appear in the Tanach, but only in the later books. In the earlier books (like Hoshea 2:3, and the ketiv of Yehoshua 2:13), the plural is אחות. He writes:

The two forms considered above, עֵדוֹת and אָחוֹת, are feminine nouns ending in –ot. This is also the common plural ending for feminine nouns. In Hebrew of the First Temple period, the form אֲחָוֹת / אַחֲוֹת was used as the plural of אָחוֹת.
…
This situation evolved further in the Second Temple period. Since consonantal waw tended to be elided between vowels, the plural form of אָחוֹת resembled the singular form (אֲחָוֹת/אַחֲוֹת > אֲחוֹת), and so dissimilation gave rise to an unambiguous alternative: *אֲחָווֹת > אֳחָיוֹת. The latter took the place of the ancient form in the later books of the Bible – the prose framework of Job and the book of Chronicles (besides an early precursor in Ezek 16:2), and then in RH.

He also explains how the vocalization of the word developed:

The plural of ‘ahot in Hebrew of the First Temple period was ‘ahawot. The written forms that the masoretes, following the pronunciation tradition of their time, vocalized, conceal forms that were sometimes pronounced differently at an earlier stage in the evolution of Hebrew. Thus, the ketiv אחותי in Josh 2:13, which stand for the plural ‘my sisters,’ should have been vocalized as אַחְוֹתַי; while ולאחותיכם in Hosh 2:3 should have been vocalized וּלְאַחְוֹתֵיכֶם.

So in ancient Hebrew, sister was achot, but sisters was achvot. To avoid the confusion, a newer pronunciation developed for sisters - achayot.

We see from this that linguistic confusion is nothing new. It is interesting though that there are those who fight strongly against changes to make language easier, when such changes were made even back in Biblical times...

hora and horim

logAfter discussing av and em, lets discuss the more generic term for mother and father: horim הורים - "parents". An informal survey I took of both native and non-native Hebrew speakers found that many of them thought that word was related to the word for teacher - moreh מורה and teaching - horaah הוראה. However, the word has a different root altogether. Moreh derives from the root ירה - "to teach", which is also the root of the word תורה - torah. Horeh comes from the root הרה - "to conceive, become pregnant". From this root we get the word herayon הריון - pregnancy.

The word is of biblical origin, but the story is a little strange here. Twice in the Tanach we find the word hora הורה - in Shir HaShirim 3:4 and Hoshea 2:7. In both instances, the word clearly means "mother". And that makes sense - the one who becomes pregnant is indeed the mother.

We do find the word horim, actually horai הורי, in one verse - Bereshit 49:26, in Yaakov's blessing to Yosef:

בִּרְכֹת אָבִיךָ, גָּבְרוּ עַל-בִּרְכֹת הוֹרַי, עַד-תַּאֲוַת, גִּבְעֹת עוֹלָם

The JPS translates this as: "The blessings of your father surpass the blessings of my ancestors [horai], to the utmost bounds of the eternal hills."

However, Sarna comments on the translation "my ancestors":

Hebrew horai is so rendered based on postbiblical usage. However, the stem h-r-h in the Bible can only mean "to become pregnant" and is, of course, solely used in the feminine. Seeing that "mountain(s)" - "hill(s)" is a fixed pair of parallel terms in Hebrew poetry, occuring more than thirty times in that order, Rashbam is undoubtedly correct in connecting horai here with har, "mountain." The Septuagint indeed reads here "ancient mountains," joining the word to the following 'ad. The phrase harere 'ad, "ancient mountains," appears in Habakkuk 3:6 in parallel with give'ot 'olam, "eternal hills." The Blessing of Moses to Joseph in Deuteronomy 33:15 employs the same imagery, though in variant form: "With the best from the ancient mountains, / And the bounty of hills immemorial..." Therefore it is best to render here, "the blessings of the ancient mountains".

I find Sarna rather convincing here. Hora certainly seems to mean mother, and the verse in Bereshit does look like it is referring to mountains. So biblically we only had a word referring to the mother, and in post-biblical Hebrew the generic term horeh for "parent" developed. The word "parent" itself in English had a similar development:

from O.Fr. parent (11c.), from L. parentem (nom. parens) "father or mother, ancestor," noun use of prp. of parere "bring forth, give birth to, produce," from PIE base *per- "to bring forth"

I haven't seen proof of this, but it wouldn't surprise me if the Latin parentem (or an earlier version) originally meant "mother", since she is the one who gives birth.

What I do find strange is how the words are listed in the dictionary. Both Klein and Even-Shoshan have listings for hora, horeh and horim. Even if they accept the more traditional understanding of Bereshit 49:26 as "my parents", I don't see why they need to have a separate entry for the plural form of the word. I suppose it's to say that the singular horeh means "father", not "parent", whereas only the plural horim is truly generic. But that's not the way it is used in Modern Hebrew. The popular online Hebrew-English dictionary Morfix translates horeh only as "parent", and doesn't have an entry for hora - as mother - at all.

Morfix does, of course, have an entry for another meaning of hora - the folk dance. And no, they're not related. As Philologos discusses here:

“Hora” comes from ancient Greek khoros, which also gives us such words as “chorus” and “choir.” Traditional circle dances deriving their names from khoros can be found all over the Balkans and southeastern Europe. They include the Turkish and Romanian hora, the Bulgarian horo, the Montenegrin and Macedonian ora, and the Russian khorovod, and they are all very old and highly similar in the way they are danced.

And the Online Etymology Dictionary provides the following background to "chorus":

from Gk. khoros "band of dancers or singers, dance, dancing ground," from PIE *ghoro-. In Attic tragedy, the khoros gave expression, between the acts, to the moral and religious sentiments evoked by the actions of the play.

And just in case I have any readers who are accustomed to the Ashkenazi kamatz - there is also no connection to the hora found in the terms "loshon hora" or "yetzer hora". I'll be sure to let you know if I stop using Israeli pronunciation...

av and em

I'd like to start a new series about kinship terms. I've written about some of them before, and you can find all of them by clicking on the label below this post, or in the categories section in the sidebar.

I think the Hebrew terms for mother - אם em, and father אב av. Klein writes that both of them "probably derive from a child's word", i.e. baby language. Similar sounds are used in many languages. In modern Hebrew we have the terms aba אבא - "dad" and ima אמא - "mom". Klein writes that they come from Aramaic, where they mean "the father" and "the mother" respectively. However, Kutscher points out that in Talmudic Hebrew aba and ima can mean "my/our father/mother". As a proof of this, he brings the story from Pesachim 4a (according to the version of Rabbeinu Chananel), where Rabbi Chiya asked Rav if "Abba" is alive. Rav didn't want to say anything offensive - so he said "Abba" is alive. Kutscher writes that Rabbi Chiya was asking if his own father was alive, and Rav knew that he wasn't. So when Rav said "Abba" was alive, he meant, "yes, my father is alive".

Because of the similar sounds in other languages, it can be difficult to determine if a word in a non-Semitic word that sounds like em or av was borrowed from Hebrew or a related language. However, there are a few examples where this seems to have happened.

For example, the Online Etymology Dictionary claims that the English word "aunt" might be connected to em:

from Anglo-Fr. aunte, from O.Fr. ante, from L. amita "paternal aunt" dim. of *amma a baby-talk or non-I.E. word for "mother" (cf. Gk. amma "mother," O.N. amma "grandmother," M.Ir. ammait "old hag," Heb. em, Arabic umm "mother").

While the origin of aunt is questionable, the words "abbot" and "abbey" definitely come from Hebrew. Kutscher, quoted here, writes as follows:

Another instructive example is the use made in Christian times of words derived from the Hebrew אב (av – “father”). In Mishnaic Hebrew we find the Aramaic אבא (abba), the earliest written form of which appears in the New Testament in Mark 14:36. The word followed hard in the wake of the spread of Christianity throughout Europe. In medieval Latin it took the form of abbas, a name for a monk, whence abbatia, a monastery, still retained in French abbaye. Furthermore, by the addition of a Greco-Roman suffix a feminine form was created, abbatissa. In English one finds abbot (and abbess), in German Abt the head of a monastery, in French abbé a priest. Likewise we have in German Äbtissin head of a nunnery by adding in, a German feminine suffix.

It is strange that the female head of an abbey would have a title deriving from the Hebrew word for "father" - but that's how language goes!

neter and nitrogen

In my last (real) post, I discussed the origin of the word soda, and its derivative sodium. Well, have you ever noticed that the chemical symbol for sodium is Na (as in NaCl - sodium chloride)? Clearly Na is not an abbreviation for "sodium", so where did it come from?

Let's discuss the development of the symbol. It might not surprise you to see that there's a connection to Hebrew here as well...

We begin with the Egyptian word ntr (I've also seen it spelled netjeri / netjry /ntrj / ntry). The word originally meant "divine" or "pure". It referred to a salt called sodium carbonate, also known as "washing soda" or "soda ash" (and sometimes simply "soda"). It was found in the area of Wadi El Natrun (apparently the wadi was named after the salt, not the other way around), in salt lakes. It was used both for soap as well as in the process of mummification.

From Egyptian the word was borrowed into Semitic languages - Akkadian nit(i)ru, Aramaic nithra, and Hebrew neter נתר. This word appears twice in the Bible - Yirimiyahu 2:22, where it talks about its use in cleaning and Mishlei 25:20, where it describes the effect of mixing it with vinegar (remember those volcano experiments from grade school where you mixed baking soda - sodium bicarbonate - and vinegar? Same thing here.)

However, in Israel, outside of Egypt, neter didn't only refer to soda (sodium carbonate) but primarily to potash (potassium carbonate), which was derived from plant ashes. (Note that we've previously discussed the Semitic origin of the symbol K for potassium, and its connection to plant ashes.) By Talmudic times (Shabbat 90a), they distinguished between the two types: Alexandrian neter, which came from the Egyptian salt lakes, and Antipatrisian neter, which came from the town of Antipatris - made from the ashes of plants.

The Greeks also had a term for this kind of soda - nitron (sometimes spelled litron). There's some debate as to whether the term was directly borrowed from Egyptian or perhaps entered Greek via Hebrew / Phoneician. From Greek the word entered Latin as nitrum and Arabic as natron.

If we continue following the word in the European languages, we see that from here it splits into two different paths. The Arabic natron enters as is into Spanish and then French, where it meant both potash and soda. The word natron is still used in English (where it appeared first in 1684), but mostly to refer to the older use of the salt.

In the 18th century, chemists began the process of identifying the various chemical elements. As discussed here, there was a split between Germanic speaking countries and the English and French speaking countries. Sir Humphry Davy was able to isolate the element sodium. As it came from "soda" - sodium was the name he chose, and this is the name used in French and English speaking countries. However, the Germanic speaking countries follow the chemists Gilbert and Klaproth who first called the element natronium (the suffix -ium means "metal", i.e. this is the metal component of soda), and then later Berzelius who came up natrium. And as we saw with Potassium and K, while the name of the element is sodium in English, the symbol of Berzelius was adopted - Na.

In Hebrew the word for sodium is natran נתרן. I haven't seen its coinage discussed extensively, but I imagine that on the one hand the influence of German science is felt here, as well as a desire to connect it to the Biblical word neter.

We've now followed the path of the word natron. However, we find another word that derived from the Greek nitron and Latin nitrum in European languages - nitre. The word is first found in French, and from there to English, where it is spelled nitre in the UK and niter in the US. Nitre does not refer to sodium or potassium carbonate, but potassium (or sodium) nitrate - better known as saltpeter (US) or saltpetre (UK). This substance isn't used as a detergent, so the translations of the Biblical neter as nitre are incorrect. Nitre was used as a fertilizer and as an explosive (for an interesting description about the connection between fertlizers and explosives, read Pollan's The Omnivore's Dilemma.)

Again in the 18th century, Rutheford discovered the chemical nitrogen, but the name was given by the French chemist Chaptal because it produces ("generates") nitre. So we actually have two elements - nitrogen and sodium, that derive from the same word.

The only question that still remains is why did nitre come to mean potassium nitrate instead of sodium carbonate? I've seen a few theories. This book (page 89) tells us that

the alchemists who first described these salts were not then able to distinguish clearly between the two types of salt.

It also points out that "nitre" stopped meaning "natron" in English in the middle of the 17th century, around the time that English began using the word natron (which must have helped with the clarity necessary to begin identifying elements like sodium and nitrogen in the following century).

This article indicates that the confusion between nitre (saltpeter) and natron (sodium carbonate) goes back very far:

It seems established that Greek nitron and Latin nitrum were used for both saltpeter and soda, which were not recognized as different substances. Pliny definitely used nitrum for both.
...
With better chemical knowledge, brought to the West by the Arabs, came also an Arabicized form of nitron, natrun, which existed beside the native Arabic word for soda, qali, meaning a plant ash, or milh qali "ash salt." As now two words were available, and two substances were distinguished, soda was called natron, while nitrum was specialized for saltpeter.

One other theory is presented by this book from 1844 (I don't know how well it has aged):

Their nitrum, however, must have been exceedingly various in its properties. For this incrustation is not always calcareous saltpetre; it is often soda, mixed with more or less calcareous earth ; and sometimes it consists of salts of sulphuric acid.
...
Substances so different ought not indeed to have been all named nitrum ; but before natural history began to be formed into a regular system, mankind in general fell into an error directly contrary to that committed at present. Objects essentially different were comprehended under one name, if they any how corresponded with each other even in things accidental. Whereas at present every variety, however small, obtains a distinct appellation; because many wish to have the pleasure, if not of forming new species, at any rate of giving new names. The elephant and rhinoceros were formerly called oxen ; the sable and ermine were named mice, and the ostrich was distinguished by the appellation of sparrow. In the like manner, calcareous saltpetre and alkali might be called nitrum. The ancients, however, gave to their nitrum some epithets, but they seem to have been used only to denote uncommon varieties.
...
But were the ancients, under the ambiguous name of nitrum acquainted with our saltpetre? There is certainly reason to think that it became known to them by lixiviating earths impregnated with salts. There are, as already said, not only in India but also in Africa, and particularly in Egypt, earths which, without the addition of ashes or potash, give real saltpetre, like that of the rubbish-hills on the road from new to old Cairo, and like the earth in some parts of Spain. It is a knowledge only of this natural kind of saltpetre, which required no artificial composition, that can be allowed to the ancients, as it does not appear by their writings that they were sufficiently versed in chemistry to prepare the artificial kind used at present.

But even admitting that they had our saltpetre, where and by what means can we be convinced of it ? Is it to be expected that any of the before-mentioned characters or properties of this salt should occur in their writings? They neither made aquafortis nor gunpowder; and they seem scarcely to have had any occasion or opportunity to discover its deflagration and the carbonization thereby effected, or, when observed, to examine and describe it. No other use of our saltpetre which could properly announce this phenomenon has yet been known. How then can it be ascertained that under the term nitrum they sometimes meant our saltpetre? Those inclined to believe too little rather than too much, who cannot be satisfied with mere conjectures or probabilities,but always require full proof, will acknowledge with me, that the first certain accounts of our saltpetre cannot be expected much before the invention of aquafortis and gunpowder. It deserves also to be remarked, that the real saltpetre, as soon as it became known, was named also nitrum ; but, by way of distinction, either sal nitrum, or sal nitri, or sal petrae. The first appellation, from which our ancestors made salniter, was occasioned by an unintelligible passage of Pliny, which I shall afterwards point out. The two other names signify like sal tartari, sal succini, a salt which was not nitrum but obtained from nitrum. Sal-nitri, therefore, or salniter, was that salt which, according to the representation of the ancients, was separated by art from nitrum, yet was essentially different from the nitrum or soda commonly in use. Birin-goccio says expressly, that the artificial nitrum, for the sake of distinction, was named, not nitrum, but sal nitrum.
...
In the course of time men became acquainted with the purer, more useful, and cheaper mineral alkali which was furnished, under the name of soda, by the Moors and inhabitants of the southern countries, who had learned the method of preparing it. The vegetable alkali also was always more and more manufactured in woody districts, as an article in great request, and sold under the name of potash, cineres clavellati. All knowledge of the impure alkali from the incrustation of walls was then lost ; and as there was no further need of guarding against confusion, it was not longer thought worth while to name saltpetre sal nitri: it was called nitrum and the oldest signification of this word being forgotten, it was admitted without further examination, that the nitrum of the ancients was nothing else than our saltpetre.

At this point the answer of when the terms diverged is not entirely clear. But I do very much like the quote above: "many wish to have the pleasure, if not of forming new species, at any rate of giving new names". I suppose if it wasn't so, I would have a lot less to write about...