Hello,

Wonder if anyone can give a brief explanation of why the word 'forc' takes an urú in the phrase "leis an bhforc" but does not with the phrase 'scian agus forc'?

Any help would be much appreciated.

Go raibh maith agaibh!

Jason

The first thing to note is that not all dialects require eclipsis (an urú) in this instance. I believe Ulster Irish would render it with lenition (a séimhiú) instead, leis an fhorc.

With that caveat out of the way, the simple explanation is that some words in Irish cause initial mutations (both eclipsis and lenition) on a following word, and others don't. The preposition i causes eclipsis on a following word, for example, i gCorcaigh "in Cork", i mBaile Átha Cliath "in Dublin", the preposition ó causes lenition on a following word, ó Chorcaigh "from Cork", and the preposition ag doesn't cause any initial mutation, ag teach "at a house". Many prepositions (but not all), when combined with the singular article an eclipse a following word, and this is why in your example leis an eclipses the f of forc. The conjunction, agus, does not cause any initial mutation on a following word, however, and that's why the phrase scian agus forc has unlenited forc.

The complicated explanation goes back into the history of the language. In Irish nouns have cases, they change depending on how they're used. Thit forc ar an talamh "a fork fell on the ground" (nominative case), but barr an fhoirc "tip of the fork" (genitive case). The dative is another case, and it typically affects nouns which follow prepositions (like leis). Modern Irish doesn't really retain the dative case, except in some fossilised expressions, for example Éire "Ireland" (nominative), in Éirinn "in Ireland" (dative), Bus Éireann "Bus of Ireland" (genitive). In most cases the nominative and dative forms of words are the same in the modern language, though this was not the case in Old Irish where nominative and dative forms were different almost all of the time. Not only did the noun itself change to be in the dative case following a preposition in Old Irish, but if the article appeared before it, that changed as well, for example, a cenn "the head" (nominative), in chinn "of the head" (genitive), lasin chiunn (dative). These articles caused different initial mutations on the following words depending what case they were in. While the different forms of the articles and most of the separative dative noun forms haven't survived into modern Irish, the initial mutations caused after prepositions, and after articles following prepositions have hung around. They are a holdover from an older form of Irish which had an entirely separate dative case.

Thank you, Ade, for that incredibly comprehensive explanation.

One quick follow up question regarding Old Irish. Any idea what the original purpose of having the different noun forms for the different cases might have been? It makes sense that most are no longer in use, but I'm not sure I understand why then existed in the first place.

The majority of languages in Europe today, as well as many in the Middle East and India, can be traced back to a single ancestral language from which they all evolved. This language, reconstructed as Proto-Indo-European (PIE), would have had these case forms and passed them on to the languages which descended from it. This is why languages like Latin and Old Irish retained many different cases.

Given the antiquity of PIE, and the fact that it predates any writing systems, it's very difficult to know much about it or how it evolved itself. It may be safe to say that it inherited case forms from an even older language from which it evolved, but when, how or why they first appeared are very difficult questions.

What we can say is that modern languages seem to lose case endings when their syntax becomes set. For example, in English, we know who is doing what by the order of words. If I say "Mary kicked John" or "the mechanic fixed the car", you know that the first person or thing mentioned (the subject) is doing the action, and the second person or thing mentioned (the object) is receiving it. Mary (the subject) is doing the kicking, John (the object) is receiving the kicking. The mechanic (the subject) is doing the fixing, the car (the object) is receiving the fixing. Because of this strict word order, we say that English is a Subject-Verb-Object language. People who speak English intuitively know who is doing actions or receiving them based on the order of the words.

This set word order is not as common in many older languages. Latin, for example, is generally described as a Subject-Object-Verb language, imagine something like "Mary John kicked", but other word orders are also very common, so for example, "John Mary kicked", "kicked John Mary", "kicked Mary John", etc. could all mean the same thing, i.e. "Mary kicked John". This freedom to play with word order is great for poetry and literature, because writers aren't restricted to using specific words in specific orders which may break rhyme schemes or meters. The downside is that it causes a lot of ambiguity. If word order isn't a determiner of who is the subject of the verb and who is the object, it's necessary to have another mechanism for determining this. This is what cases were useful for. If the names, Mary or John, changed form depending whether they were the subject (hence, nominative case) or object (hence, accusative case) of a verb, then it would be clear who was doing the action and who was receiving it regardless of the word order.

Without knowing the ultimate origins of features like word order or varying grammatical word forms in very archaic or prehistoric languages, what we can say is that some of the earliest written languages we have, which are descended from PIE, demonstrate that syntax (word order) was not relied upon as heavily as it is in many modern languages, also descended from PIE, to determine semantics (the meaning of a word, or words used in combination). Instead many older languages required case endings to help carry the same meaning where free word order couldn't. In modern languages there seems to have been a shift towards strict syntax to encode semantics, making case forms somewhat redundant. The result is that many of them either have fallen out of use or have been simplified in modern languages.

English has the articles the and a. My native language works perfectly well without them, they’re redundant. It makes sense that in many languages there is no such thing, so why do they exist in English?

The answer is: because different languages encode information in different ways. Polish doesn’t require the articles as with its freer word order there are other ways of encoding what the new information (the rheme or focus) of discourse is and what is the context (the theme or topic). English with its strict syntax needs some other way of conveying the same information, and employs the article to communicate what is already known among the speakers and listeners and what is just being mentioned for the first time.

Similarly noun cases, in languages that make heavy use of them, encode a lot of different relations between ideas which in English are most often encoded using prepositions (Tom’s house – using the genitive case, or at least one of the last remnants of it in English = the house of Tom, using the preposition of; Give me the book using the “ditransitive” construction that continues Old English dative case, me here continues the dative form meaning “to me” = Give the book to me using the preposition to.

Languages have different structures and the fact that something doesn’t work the same way in English doesn’t mean the same thing makes sense in other languages. Also the fact that today in a given language something is one way doesn’t mean it was the same a few centuries before.


Some of the case endings (especially dative, instrumental, ablative; at least plural), if not the cases themselves, might have been innovated a bit later, during the Proto-Indo-European stage or even in the daughter branches separately after some early split (as we see quite a bit variation in the endings). But, sure, can’t say much beyond mere speculation without more data.



This is extremely western European-centric view. There is nothing “old” in having a freer word order and nothing “modern” about losing them. Also nothing “simplified” in losing the cases – while the morphology is simpler, the stricter syntax isn’t. Languages change, in their morphology and syntax, they don’t “simplify”. Baltic languages have innovated a few new cases, probably under Finnic influence, since the early Middle Ages – instead of losing them. Slavic languages innovated new accusative for animate masculine nouns – making the system more complex. While most languages in the west of Europe lost most of their nominal morphology – but that’s not a general rule. Tocharian, an old IE language – but attested long after its separation from other IE languages – also innovated a good set of new cases, rather than losing them.

There is no set trajectory from inflectional morphosyntax to analytical one with set word order. Languages with rich morphology and freer syntax are in no way “less modern” or “older” from analytical languages. Whatever most Celtic, Romance, and Germanic languages, including English, did in the last 1000 years.

But thanks for implying my native language isn’t modern.

For me, Irish absolutely does have four cases. The example given above of "ag teach" strikes me as functionally illiterate. I think this may describe what has happened to Irish in large parts of the Gaelthacht, but it is absolutely correct to say and write "ag tigh". Caraid, coin, teagain, the slenderisation of feminine nouns in the dative singular is absolutely correct literary Irish.

You're right, I was answering the question in the context of Irish, and making comparisons to other languages with which I am familiar. I'm no expert in Baltic or Slavic languages, and I completely accept that there are blind spots in my knowledge of the development of other IE languages.



Sorry if I touched a nerve, silmeth. I think you know, though, that I was neither suggesting your language, nor any other modern language which happens to have case forms, is somehow antiquated or simplistic. Even Irish maintains nominative, vocative and genitive cases, and as djwebb has pointed out above, there is a strong argument to be made for the dative. So, if I were insinuating that about your native language, I'd have been insinuating it about my own ancestral one also. I never claimed it was a general rule that languages lose cases and develop stricter syntax. If you look again at what I said you'll see I was quite specifically suggesting this tendency only in languages which do develop a stricter syntax:



Similarly, I never suggested that historical languages can't display stricter syntax and fewer case forms. I just observed that "many" older languages had case forms and freer word order than languages which developed from them:



You're right, of course, to point out that a stricter syntax makes up in complexity for loss of cases. Still, the case system absolutely is "simplified" where the number of cases are reduced. The language as a whole maintains complexity, it just shifts away from the case system. Nothing in what I said suggested that languages as a whole "simplify" over time, or that languages, modern or historical, can't innovate new case forms.

In any event, I think my assessment still stands in the context of Irish, at least. The case system has been simplified over time as the language developed. There are the remnants of a dative case, and the accusative is gone.

Actually the situation in Irish is part of the reason I’m sensitive to it. Very often features of learners’ Irish with heavy English influence are perceived as “more modern” due to their similarity with English than “old, outdated” Gaeltacht speech – mostly popular sentiment, but sometimes also expressed in sociolinguistic papers. And I think it’s really harmful towards a minoritized culture. So I feel the need to voice my opposition whenever I see any discussion about languages using the “modern (ie. English-like) vs old (ie. more exotic from English perspective)” framing.


That’s fair, although I’d argue that the causal chain is often the reverse – first the case endings were dropped in the west due to apocope, which made it difficult to keep freer syntax – since there is no case agreement between words in a sentence, it’s more difficult to unambiguously communicate the relations between them, and more difficult to communicate their roles in the sentence – which then required more stricter syntax (with no breaking of constituents, and fixed places of subjects and direct objects, etc.). So I think it’s rather that languages that lose the cases develop stricter syntax. But the effect is the same.

And then, Old Irish already had fairly strict syntax while still keeping the cases, so there’s probably still more to it and I’m not quite right either.

It is the case, however, that synthetic languages tend to become more analytical over time, rather than the other way round. The example of Tokharian is a lame red herring. Not that much is known about the language compared to our knowledge of modern languages, and it is thought that the development of agglutinative cases came under influence of Samoyed languages they came into contact with.

If you see this thread at https://www.reddit.com/r/linguistics/co ... _language/ , you will notice that in Latin, Sanskrit, Koranic Arabic etc, there is a tendency for older languages to be more morphologically complex. Yes, it is true that loss of morphological complexity can result in more complex syntax and word order - I think L2 speakers of English frequently imagine that English is a simple language, not really realising in many cases how badly they speak the "simple" language.

It would, however, be amazing for English to suddenly develop a dative case or an instrumental case. In fact, although I pointed out that the dative case in Irish is used in some areas, or particularly in some words and phrases, it is undeniable that the trend has been away from the use of morphologically apparent cases. Clearly, loss of all the cases is the trend in the language. Maybe the genitive singular is the one case that will survive? Ó Siadhail argued there are two vocative plurals and seven genitive plurals (if I recall correctly) in Cois Fhairrge. In fact, when people say "you and I" instead of "you and me" in English, even in cases where the whole phrase is in the object position, they are showing some tendency towards weakening of the object forms in English too.

Irish has cases - or traditional Munster Irish had, them, let's put it that way - but unlike, eg Russian, where the case alone shows the relationship, eg другу, "to the friend", is a case that encapsulates the whole meaning, the Irish don charaid manages to be both synthetic and analytical at the same time. Analytical because the "don" is required there. Caraid on its own can't give the meaning "to the friend". Synthetic, because the noun was declined for the dative too. You can see how easily don charaid degrades to don chara (although a nod has to be given to the use of the dative for the nominative in Cois Fhairrge and elsewhere, so that it is caraid in the base form anyway, and so "don charaid" too).

I think there are signs in Russian of a small move towards analytical forms. There are phrases where in the 19th century the straight dative was used, whereas now a preposition would be used too: другу>к другу. I could find examples in my files, but it isn't on topic in an Irish forum. Maybe there are examples in Polish too. In fact there is a PDF by a Pole on the growing use of analytical forms in Polish at https://cejsh.icm.edu.pl/cejsh/element/ ... 5-1162.pdf

I absolutely share your concern for minority languages in that regard. Funnily enough, I lived with a native Connemara Irish speaker and his wife for two years before Covid. They used to address each other as mo ghrá, and said they found it archaic to use the vocative. They'd look at me funny when I'd use it and say it's a feature of learners' Irish. Granted, this is a sample size of one I'm dealing with.



Sure, that might well be the case. I'd have been pretty much as happy to have written "What we can say is that modern languages' syntax seems to become set when they lose case endings." My main aim was to outline the correlation, not to infer causation in one direction or another, but this can be hard to avoid.