Why do some conquerors replace the language and some not?

[Owain]

Putting this here because it's mostly about trying to work this out for conworlding purposes, and is only indirectly linguistic really.
For instance, Latin obviously replaced local languages in what's now France, Iberia, and part of the Balkans (and the rest of Italy, I guess), Saxon replaced presumably both Latin and Brythonic in Britain, Spanish replaced various local languages, and similar things happened in the Scramble for Africa and the British conquest of India, to greater or lesser extents (I'm largely ignoring cases like the US because clearly there it's to do with wholesale replacement of the population more than getting conquered peoples to speak your language). But Frankish went nowhere in replacing Latin/French in France, the same with Gothic in Spain and Italy, or French in England.
Or for a case with multiple examples of both in the same area, why did Akkadian and later Aramaic (and later still Arabic) seem to have established themselves as vernaculars in central and southern Mesopotamia, but not Assyrian or Persian or whatever the Kassites spoke? Particularly since the Arameans were actually one of the shortest-lived conquests from what I can tell?
The two explanations I've been able to come up with from looking up different cases are urbanism and administration. If you require your language to be used for administration the conquered peoples are going to need it to be at all successful - perhaps ironically, this is probably worst for the local language when there's some possibility for local people to become administrators, so long as they speak the conquerors' language, because then it's something to aspire to for better things rather than just keeping more of what you've got. And this tends to associate with the examples above (relatedly, I'd assume Arabic was probably helped along by the conversion-by-tax-policy approach in a similar way to how Islam was), except that I know in England after the Norman Conquest, legal documents were in Latin and legal speech was in French (I think the latter wasn't technically abolished as the default till the 1700s, but obviously fell into disuse in practise a lot earlier).
The other is that if a conqueror is establishing new colonial cities - or sacking existing ones and re-establishing them according to their own design - it's especially easy to ensure these new towns speak the language you want them to, and in general languages tend to spread from cities to rural areas rather than the other way round. Again this generally seems to correlate with the above, but with counterexamples - one that particularly springs to mind for me is that round here in Wales, while both have happened, the big change was a fairly rapid one long after it had begun (although I guess associated with more of a step up on the administrative aspects, and in line with universal schooling, which I haven't really considered), and often the earlier towns originally intended to plant English in Wales are more Welsh speaking than the ones founded later with less of a deliberate policy of not having Welsh spoken in them.
So between them they seem to explain most of them, but it all seems fairly random what happens when only one applies, and in particular I'm not sure either really explain Aramaic, which seems to have done most of its spreading after the conqueror/conquered situation had actually reversed in northern Mesopotamia, if not so much in the centre and south.
So does anyone have any thoughts?

[mèþru]

Assyrians were an Akkadian peoples
Actually, many indigenous American languages have died out with their ethnic groups still surviving to the present.

In the case of Anglo-Saxon, Latin and Arabic the local peoples heavily intermarried with and were assimilated into the larger ethnicity: several Semitic peoples who didn't assimilate speak modern descendants of Aramaic. Gaulish was still spoken in some parts of rural France up to the 9th or 10th century; Basque is still alive today. In France, Latin survived because of several reasons that didn't apply to England: the Franks converted to Christianity early on (so Latin had prestige among them), they were heavily outnumbered by local peoples and the country was much more Latinised at time of conquest.

[Owain]

Assyrians were an Akkadian peoples
Actually, many indigenous American languages have died out with their ethnic groups still surviving to the present.

In the case of Anglo-Saxon, Latin and Arabic the local peoples heavily intermarried with and were assimilated into the larger ethnicity: several Semitic peoples who didn't assimilate speak modern descendants of Aramaic. Gaulish was still spoken in some parts of rural France up to the 9th or 10th century; Basque is still alive today. In France, Latin survived because of several reasons that didn't apply to England: the Franks converted to Christianity early on (so Latin had prestige among them), they were heavily outnumbered by local peoples and the country was much more Latinised at time of conquest.

Yeah, I was oversimplifying in that the southern language still seems to generally get called Akkadian after they separate.
I guess similarly what I mean is how a minority language becomes a majority one without its speakers at the starting point growing larger in the genepool as it were rather than how I phrased it initially.
The intermarriage point is a strong one (and I think most of the surviving Semitic languages are in communities that didn't convert as well, although not vice versa?), but to an extent raises the question of why it had that effect and not the opposite one (which I think it did to at least some extent in the case of the Norman Conquest), especially as I would have thought at least at first it would generally be more conquering men marrying local women, and we know mothers tend to have more influence than fathers on language transmission.

[mèþru]

It also depends on the situation. In Arab culture, the son of an Arab man is considered an Arab.

In the Norman conquest, the Normans were few and had no agenda to culturally convert the natives. In fact, their claim was based off William being related to a previous ruler. They wanted to show that they are in fact rulers by right of native law.

[Frislander]

Part of the thing with Aramaic was that it was actually the main administrative lingua-franca of the Achaemenid Persian Empire, I think mainly because the ruling Persians didn't want their language to be spoken widely by the common people.

[mèþru]

The Persian empires didn't care about extinguishing local cultures. Rome thought that "civilising" "barbaric peoples" would make them easier to administrate. They didn't attempt much to Romanise the Greeks, who they saw as a culture worth emulating. They Romanised fellow Italic peoples because their own identities had repeated led to rebellions against Latin rule. Arabic had to be spread as the language of the Koran. Each culture had their own underlying reasons that must be explored on a case by case basis. I don't think general models are really that useful.

[Zaarin]

Part of the thing with Aramaic was that it was actually the main administrative lingua-franca of the Achaemenid Persian Empire, I think mainly because the ruling Persians didn't want their language to be spoken widely by the common people.

The Persians didn't care about keeping their language a secret. Aramaic had been the lingua franca of the Assyrian Empire since Tiglath-Pileser III's conquest of Aram; the Persians simply made its use official. The Persians were notably liberal in permitting conquered peoples to preserve their local cultures and identities; adopting the pre-existing lingua franca was a pragmatic choice. NB the Achaemenid kings continued to make inscriptions in their homeland in Old Persian and continued to pass on their religious rites in Avestan.

As for speculating why the Kassites abandoned their language in favor of Sumerian, the Sumero-Akkadian cultural complex in Mesopotamia was a cultural force to be reckoned with. It's hardly the only instance of a "barbarian" dynasty adopting the native culture in the region (cf. the Hyksos, Canaanite, Nubian, and Libyan dynasties in Egypt, all of whom assimilated to Egyptian culture).

Also, as mèþru pointed out, Aramaic is still spoken in the Middle East, chiefly by Christian ethnoreligious minorities like the Assyrians (but also by some Jewish communities and the Mandaeans).

[zompist]

A lot of it comes down to numbers. The Normans were a tiny fraction of the English population. For that matter, the Persians were very few compared to the inhabitants of Mesopotamia and Anatolia.

As for the Germanic invaders of Rome, we have one firm number on this: Gaiseric had his Vandals counted in 429, presumably to estimate shipping costs as they were invading North Africa. The total including women and children was 80,000. Compare to the total population of Italy, at least 6 million.

Some of it must be cultural. E.g. a writing system for Persian was only devised during the empire, and only used for inscriptions; it was impossible to make it the language for administration or diplomacy or literature. As mentioned, instruction in Arabic is important for Muslims, while (e.g.) Christians are mostly happy with translation.

Some of these conquests were far less thorough than we might think. E.g. the Spanish didn't obliterate Quechua, which was the main language of the native (non-mestizo) population (and still is, in southern Peru). Arabic didn't overrun the conquered people's languages in the Maghreb, Spain, Anatolia, Persia, or India (though of course it was the source of a load of borrowings).

The real mystery might be why the Romans did so well in Spain and Gaul. Contrary to our usual idea of them, they didn't succeed that well at urbanization— the cities of the West collapsed when the empire fell. Still, they probably indicate a high number of immigration from Italy, and the settling of a large (Vulgar) Latin-speaking Roman army on the frontier must have been a factor.

It's also pretty amazing that Arabic took over Egypt. Part of this may simply be that the process took a very long time. And perhaps Egypt, which hadn't been ruled (for long periods) by native Egyptians for 1200 years, no longer had a strong native identity.

[zompist]

NB the Achaemenid kings continued [...] to pass on their religious rites in Avestan.

Just as a note— I don't think we actually know this. I've been reading histories of Persia, and it's remarkable how much we don't know! We know that Darius was devoted to Ahuramazda— but it's not even certain that this is identical to what we know as Zoroastrianism. The Magi weren't even (apparently) Persians, but Medes... or maybe just Iranians in general; Avestan itself seems to originate in northeastern Iran or Central Asia. Parts of the Avesta are ancient, but the surviving parts seem to derive from a Sassanid compilation (centuries after the Achaemenids).

[mèþru]

I think parts of northern France were, for a brief time, heavily Frankish speaking.

Also, Arabic did eventually completely replace African Latin (well partially, as I consider Sardinian to be a remnant) and probably several Berber languages.

[Salmoneus]

To reiterate Zompist: most of the time, shear numbers are key. Language shift is almost always the result of large-scale demic diffusion, and large-scale demic diffusion almost always results in language shift.

The other two consistent factors I'd say are cultural power and political power.

Where one culture is more advanced and civilised than another, it's likely that its language will have greater influence - people attempting to emulate the more advanced culture tend to be receptive to its language, while people are less likely to adopt the language of people they see as 'barbarians'. However, it's very difficult to separate this factor out from population, most of the time, since population both leads to, and is the result of, cultural advancement. It's easy to point to Greek under Rome, for example - but the other big difference between Greek and Gaulish, other than Socrates, is that Greece was the most densely-populated part of Europe, and that Greek-speaking cities were among the largest in the world.

Regarding political power: weak rulers are likely to rule by cooption, retaining much of the culture and power structure of their conquered peoples. Strong rulers are more likely to be able to impose direct rule in their own language. But again, this is likely to be closely tied to population: small ruling elites are inherently weaker than large ones.


To these regular factors, I'd add a fourth: the formation of lingua francas. I'm not sure there's any way to predict this - sometimes, for the purposes of trade, where a linguistic vacuum exists (i.e. where economics encourages trade, but there is no political system to impose a common language), a language happens to be adopted as a lingua franca. Sometimes this is superficial, and sometimes it becomes entrenched. The chosen language will not be something small and unheard of, but it needn't necessarily be the most spoken or most powerful language either. Indeed, I think sometimes lingua franca selection has favoured second-rank languages specifically in order to avoid association with the dominant culture - though I can't think of a real example off hand. [Like, if the UN during the Cold War had been forced to choose a single official language, it might well have been, say, French, rather than either English or Russian].


There are some cases where the result has been unexpected - the triumph of Arabic in North Africa being the most obvious, probably. I think several factors are involved here - the cultural prestige of Arabic as the language of a major religion (unfortunately there aren't many, or any, analogous cases to compare to*), the political success of the Caliphate, and also I suspec the lingua franca dimension. It's notable that both the areas where Arabic surprisingly succeeded in north africa - Egypt and the Maghreb - were areas where multiple layers of language were still spoken (Egypt had Coptic, Greek and Latin, plus a multicultural urban culture; the Maghreb had Punic, Berber, Latin, Germanic, and possibly still small numbers of Iranian speakers). It's possible that in this circumstance there was a great advantage to adopting Arabic not only as the language of the government and the religion, but also simply as a common tongue to bridge a divided community. Which I suppose also suggests a further factor - influence beyond the region. Languages that give access to a wider world will probably do better than languages where the speakers have all migrated and there are no speakers left outside.

And, of course, duration.


*we might compare the Christian languages - Greek, Aramaic and Latin. But Latin's dominance was primarily political-demographic, not religious. Whereas Aramaic, and later Greek after the decline of the Empire, and particularly in areas like Russia, lacked the political dimension. Arabic is an unusual case study where there was religious and political domination, but not demographic domination.

[evmdbm]

Not much to add, but there's a very good book called Empires of the Word by Nicholas Ostler if you want to know more, explaining what makes a "prestige" language.

[Zaarin]

NB the Achaemenid kings continued [...] to pass on their religious rites in Avestan.

Just as a note— I don't think we actually know this. I've been reading histories of Persia, and it's remarkable how much we don't know! We know that Darius was devoted to Ahuramazda— but it's not even certain that this is identical to what we know as Zoroastrianism. The Magi weren't even (apparently) Persians, but Medes... or maybe just Iranians in general; Avestan itself seems to originate in northeastern Iran or Central Asia. Parts of the Avesta are ancient, but the surviving parts seem to derive from a Sassanid compilation (centuries after the Achaemenids).

Oh, I'm aware that the evidence for "proper" Zoroastrianism in Achaemenid Persia is remarkably scant. But since portions of the Avesta are certainly contemporary with Old Persian, someone in Greater Persia was passing along the Avesta until it could be compiled by the Sassanids (who were much more explicitly Zoroastrian, though it's my understanding that Zurvanism is generally regarded as heretical by modern Zoroastrians).

[Owain]

snip

snip

Thanks for the expansive answers.

Some of it must be cultural. E.g. a writing system for Persian was only devised during the empire, and only used for inscriptions; it was impossible to make it the language for administration or diplomacy or literature. As mentioned, instruction in Arabic is important for Muslims, while (e.g.) Christians are mostly happy with translation.

Related to this I guess (and perhaps also Salmoneus's point about lingua francas), someone elsewhere was suggesting that one reason for Aramaic's adoption by the Assyrian bureaucracy was to do with its script being easier for writing on paper etc than cuneiform (and I guess once it was co-official being an abjad rather than a mixed syllabary and logography might be an advantage as well). Presumably both might also apply to usage as a language for trade.

[mèþru]

That sounds like BS to me. Even when exposed to alphabetic and syllabary systems, most Chinese character using countries continued using Chinese characters until the 19th/20th centuries or never stopped.

The idea that alphabetic languages are better for trade sounds like lingering Eurocentrism in historical studies.

[Raholeun]

That sounds like BS to me. Even when exposed to alphabetic and syllabary systems, most Chinese character using countries continued using Chinese characters until the 19th/20th centuries or never stopped.

The idea that alphabetic languages are better for trade sounds like lingering Eurocentrism in historical studies.

What countries use Chinese characters exclusively? Besides Sinitic speaking peoples, I can only think of the Vietnamese who indeed abandon it "recently". But even in scripts that superficially seem similar like Tangut, Khitan and Jurchen, the number of strokes typically decreased. When trading with peoples who have a radically different type of script as an alphabet is, the alphabetic scribes are at a disadvantage as they need to learn many more signs than the 26 hypothetical Chinese traders would need to learn.

[Ketsuban]

What countries use Chinese characters exclusively?

Japan. Hiragana and katakana are based on the practice of manyōgana (using kanji for their sound to write native Japanese words) with hiragana being handwritten forms and katakana being graphic sections of more complex kanji.

[mèþru]

Also I never specified exclusivity.

[Vijay]

I thought the reason why Chinese characters are widely used in East Asian countries to this day was because of their historical and cultural prestige...Is that irrelevant here?