Reanalysis and extension also apply to semantic change, and in this post, I pick up from my second last post on Vasilis Orfanos' analysis of the semantic transition, over at the Magnificent Nikos Sarantakos' Blog. The debate that broke out between commenters on when the changeover happened is of particular interest, since it highlights the mechanism of how gradual the semantic change is.
To start with, what actually happened to "nation of Muhammad" to end up meaning "come hell or high water" is pretty clear, and I will quote Orfanos' analysis:
In Turkish there is an expression ümmet-i Muhammed, "nation of Muhammad", with which Muslims refer to the sum of their coreligionists. During Ottoman rule, it was also used as a warcry, which was interpreted by Greeks as an oath/commitment to attain victory. So it passed into Greek with the sense "at any sacrifice", as αμέτι μουχαμέτι—possibly through the influence of a folk etymology from the name Ahmed or the oath Μα το Μουχαμέτη "By Muhammad!"
Let's walk through the examples again, from a semantic rather than syntactic point this time. Unsurprisingly, the semantic and the syntactic development aligned closely.
To begin with, the expression refers literally to the Muslim nation (1, 5):
- Τον κύριο παρακάλεσαν να κάμη μερχαμέτι, Και να τους κάμη ολουνούς του Μουχαμέτ’ ουμμέτι "They begged the Lord to show compassion, and to make them all Muhammad's nation".
When the word acts as a war cry (as a vocative and then an exclamation), it expresses encouragement to action (4, 6, 7, 8, 9, 11, 13)—as well as its polar opposite, despair (3, 14, 15, 16):
- ιμέτι, Μωχαμέτη, ’Σ τον ντιν ντουσμάνη σήμερα να κάμωμεν γαϊρέτι, "Nation of Muhammad! Today against the infidel enemy we shall show endurance."
- Ορίστε, λεγ’ Αλήπασας, ιμέτι Μουχαμέτη, Χαΐρι δεν εχούμε ’μεις εφέτ’ απ’ το ντουβλέτι, "'See,' Ali Pasha says, 'By the Nation of Muhammad! We're not seeing any joy from the government this year!'"
Analysed at that level, the distinction between vocative and exclamation is not that relevant. The reanalysis is clearly when the expression is still a vocative—going from referring to Muslim troops being addressed, to an exclamation. The extension is, again, when the expression of encouragement is used without anyone explicitly being addressed; admittedly, that is clear in the expression of despair rather than the warcry.
The critical switch syntactically was when the expression was used in indirect speech: that allowed it to be reanalysed from an exclamation to an adverb (2, 10). The accompanying semantic shift is earlier: it is already inherent in a warcry, used in direct speech, that the exclamation indicates a commitment to do something, particularly in that it was followed by a command to do it:
- (6) Αμέτη, Μωαμέτη! Πιάστε τον τόπον δυνατά, "Nation of Muhammad! Fortify the place boldly."
The conflation with the oath μα το Μουχαμέτη "By Muhammad!" would have reinforced that notion of commitment. The use of the expression in indirect speech is an extension, rather than a reanalysis of that meaning of commitment: it is now separated from any notion of directly addressing people. Orfanos accordingly identified the song on Messolonghi (10) as critical to the development of αμέτι μουχαμέτι—"as an oath committing the besiegers to achieve their goals":
- Όλοι τους ωρκισθήκανε αμέτι Μουχαμέτη, στο Μεσολόγγι να εμβούν, να κάμουν κιαμέτι "They all swore, 'Nation of Muhammad!', to enter Messolonghi and cause havoc."
But αμέτι Μουχαμέτη does not actually mean anything intrinsically different in (6) and (10): the change is syntactic, not semantic. The essential meaning is still the same: "the subject is committing themselves to doing X".
From "They swear they will do X", it is an easy step to "they will definitely do X", "they will bloodymindedly do X". That is the shift in (12), where the modern use of the expression is obvious:
- αλλά ο Γαρδικιότης είναι Αμέτ Μουαμέτ κατά του στραβού, "but Gardikiotis is bloodymindedly against the blind man."
Again, the change here is syntactic: it has broadened the contexts in which the expression shows up. There is a slight change in meaning in (12), which is metonymic. X committing to doing Y contains the necessary implication that X feels strongly about Y. The semantic shift moves in (12) from commitment to an action, to indication of the associated emotion: Gardikiotis is not committing to doing anything as bloodymindedly as an Ottoman warrior: he just shows the same singlemindedness as an Ottoman warrior.
But this generalisation in meaning hasn't been followed since: αμέτι μουχαμέτι is still associated with intended action, rather than emotional state. What has been picked up by the expression, rather, is connotation, which leads to the meaning being strengthened.
The context is key to how the connotations took root; context, after all, is where connotations come from. Greeks who did not understand Turkish heard the cry in battle, preceding an undertaking to wreak havoc: "Ümmet-i Muhammed! We will burn them down!" The cry sounds something like "By Muhammad!", which cements the notion that it expresses a religiously fervent, bloodyminded commitment to do ill, made by hostile warriors. When the cry is reanalysed as a more generic adverb indicating commitment, those connotations come along with it: it has picked up the negative connotations of "being on the warpath", of unreasonableness, of disproportionate singlemindedness.
That, I hope, is revealing. The real change in αμέτι μουχαμέτι is syntactic, not semantic; the essential semantic reanalysis, from vocative to expression of commitment, is earlier (indeed, it happened in Turkish and not Greek), and it is not particularly drastic. Connotation, rather than semantic shift, has given the expression the force it now has; and connotation is a more subtle change than the syntactic reanalyses undergone by αμέτι μουχαμέτι.
Because the semantic change is not particularly pronounced, my godfather Tasos Kaplanis thought there was still a missing link between the 1820s songs with warcries (10) and the 1859 Modern Greek adverbial use (12):
What is impressive is that in most cases from the 19th century recorded in the post, the phrase has or could have the initial meaning it has in Turkish—until we reach Xenos (12) and Papdiamantis (17, 18), who record the phrase with more or less its modern meaning. […] I find this somewhat peculiar, and I still can't see the link between the popular records of the phrase (not just in folk song and older texts, but also Kolokotronis and Koutsonikas), and the literary or learnèd usage by Xenos and Papadiamantis.
In other words, in (10) the expression is still a Muslim warcry; in (12) it is a Greek adverb; and the transition between the two is not obvious.
I think the ensuing exchange is illuminating:
- Nikos Sarantakos thinks the songs of Messolonghi (10) and Xopateras (11) do provide the missing link: "they swore, amet moukhamet, they would conquer Messolonghi / they would seize the priest."
- Tasos thinks that (11) merely quotes the interjection. In fact, that's also how I'd interpreted it last post. As to the critical example (10), Tasos says:
In the siege of Messolonghi, Legrand's misunderstanding ("By Muhammad!") could be a missing link—though I still don't see how we got from "By Muhammad!" to "stubbornly, wilfully, at any sacrifice", etc. And if you add a comma to edition, αμέτι μουχαμέτι turns back into ümmet-i Muhammed: "They all swore, Nation of Muhammed, they would enter Messolonghi."
- Maria retorts she too can see the transition clearly.
- Tasos answers: "OK, Maria, maybe I can't see it because I understand what the Turkish means."
Tasos wasn't saying that because he took offence, but because understanding Turkish really does get in the way of seeing how the meaning change happened: that's why I'm highlighting the exchange. Greek speakers clearly were starting not to understand what ümmet-i Muhammed means: that's what allowed them to distort ümmet as αμέτι. Greek speakers were starting to guess what αμέτι μουχαμέτι meant from context. That meant that the meaning shift was possible as a reanalysis, even if superficially the phrases looked the same: the meaning shift happened in Greek-speakers' heads, and had not yet undergone extension to new contexts. So if you actually do know Turkish, you will not initially notice a change at all.
It's not surprising that Sarantakos, like Orfanos, highlights (10) as a pivot: it displays a clear syntactic reanalysis, rather than the more subtle semantic shift, through indirect speech. Tasos retorts that the difference between "at any cost" and "Nation of Muhammad!" in (10) is merely a comma. But of course, that's precisely why (10) matters: syntactic reanalysis is exactly a matter of a missing comma—that is, a reinterpretation of the syntactic structure, which can be represented through intonation, and in print through commas.
Sarantakos also highlights (11) as a comparable instance. I said that in the last post I interpreted (11) like Tasos, with a disjoint exclamation: "'Glory to Allah', they cried, and 'Ameti Moukhameti': they would go seize the priest, to calm Crete down." The expression does look like (10), and could be construed in the same way, as indirect speech containing an exclamation: "and (Nation of Muhammad!) they would go seize the priest" > "and at any cost, they would go seize the priest".
I don't think that's what the composer of the song intended—because "Nation of Muhammed!" was conjoined with "Glory to Allah!" But that construal was clearly starting to be possible—and in a couple of decades, led to the unambiguous extension in (12). And I hope I've explained above how to get from "By Muhammad" to "at any sacrifice", through hostile connotation.