Mutual Intelligiblility in the Romance Family (Reading)

Just a personal anecdote. I have been reading a lot of Italian lately (with the help of Google Translate). I already read Spanish fairly well. I have studied French, Portuguese and Italian, and I can read Portuguese and French to some extent, Portuguese better than French.
But I confess that I am quite lost with Italian. This is worse than French and worse than Portuguese. A couple weeks of wading through this stuff hasn’t made me understand it any better.
Portuguese and Galician are said to be so close that they are a single language. I don’t agree with that at all, but they are very close, much closer to Spanish and Portuguese. Intelligibility may be on the order of 80-90%.
Nevertheless, the other day I tried to read a journal article on Galician. It looked like it was written in Portuguese, and who would write in Galician anyway? I copied the whole thing into Google Translate and let it ride. I waded through the whole article, and I must say it was a disaster. I had a very hard time understanding many of the main points of the article.
Then I remembered that Translate works on Galician now, so I decided on an off chance that the guy may have written the piece in Galician for some nutty reason. I ran it through Translate using Galician as target. The article went through perfectly. You could understand the whole thing. It was then that I realized how far apart Portuguese and Galician really are.
You can try some other experiments.
Occitan is said to be nearly intelligible with Spanish or maybe even French, better if you know both. There’s no Google Translate for Occitan yet, but I had to deal with a lot of Occitan texts recently. I couldn’t make heads or tails of them despite by Romance reading background. So I tried using Translate to turn them into Spanish or French. French was a total wreck, and there was no point even bothering with that. Spanish was much better, but even that was a serious mess.
Now we come to the crux. Catalan and Occitan are said to be so close that they are nearly one language. Translate now works in Catalan. So I ran the Occitan texts through Translate using Catalan. The result was a serious mess, but you could at least understand some of what the Occitan texts were about. But no way on Earth were those the same languages.
People keep saying that if you can read Spanish, you can read Portuguese. It’s not true, but you can see why people say it. Try this. Take a Spanish text and run it through Translate using the Portuguese filter. Now take a Portuguese text and run it through Translate using the Spanish filter. See what a mess you end up with!
Despite the fact that I can read Spanish pretty well, I have tried to read texts in Aragonese, Asturian, Extremaduran, Leonese and Mirandese. These are so close that some even say that they are dialects of Spanish. But even if you read Spanish, you can’t really read any of those languages, and they are all separate languages, I assure you. Sure, you get some of it, but not enough, and it’s a very frustrating experience.
There are texts on the Net in something called Churro or Xurro. It’s a Valencian-Aragonese transitional dialect spoken around Teruel in Aragon in Spain. It also has a lot of Old Castillian and a ton of regular Castillian in it. Wikipedia will tell you it’s a Spanish dialect. Running it through both the Spanish and Catalan filters didn’t work and ended up with train wrecks. I doubt if Xurro is a dialect of either Catalan or Spanish. It’s probably a separate language.
There is another odd lect spoken in the same region called Chappurriau. It is spoken in Aguaviva in Teruel in the Franca Strip. The Catalans say these people speak Catalan, but the speakers say that their language is not Catalan. Intelligibility with Catalan is said to be good. So effectively this is a Catalan dialect.
I found some Chappurriau texts on the Net and ran them through Translate using Catalan as the output. The result was an unreadable disaster, and I couldn’t really figure out what they were saying. Then I tried the Spanish filter, and that was even worse. I am starting to think that maybe Chappurriau is a separate language as its speakers say and not a Catalan dialect after all.
I conclude that the ability to cross read across the Romance languages is much exaggerated.
Not only that, but many Romance microlanguages, transitional dialects and lects that are supposedly dialects of larger languages may actually be separate languages.

Please follow and like us:
Tweet 20

10 thoughts on “Mutual Intelligiblility in the Romance Family (Reading)”

  1. French uses an idiosyncratic spelling and has strange vowels(the langd’oil anyway) probably derived fro Gaulish and Frankish. Maybe this is part of the problem with google translate A similar effect from spelling makes Dutch and Frisian seem much more different in writing from even basic/rustic/archaic varieties of English with low levels of Norman/Latin, than it is.

  2. correction:
    “A similar effect from spelling makes Dutch and Frisian seem much more different in writing from even basic/rustic/archaic varieties of English with low levels of Norman/Latin, than they are”.

  3. Dear Robert
    Your argument is totally unconvincing. Instead of using a translation machine, what you should have done is to show a text in Galician to a literate Portuguese-speaker and a text in Occitan to a literate Catalán-speaker, and vice-versa.
    To understand a text, you often have to understand every word in it. If you don’t get one word, you may miss essential information. Let’s take this sentence: “I put your suitcase in the attic”. If a foreigner understands every word in that sentence except the word attic, he still can’t find his suitcase.
    Regards. James

  4. Dear Robert
    As you know, I can read Portuguese. I went to Wikipedia and read the article about Galicia in Galego. In understood everything. It was the first time ever that I read something in Galego.
    When I read Friesian, I miss a lot. Conclusion: written Portuguese and Galego are closer to each other than written Dutch and Friesian are to each other.
    Cheers. James

  5. Now I went to read the article about Galego in Catalán. That was much harder, although I can read French and Spanish. In many instances I really had to guess by relying on context. James

  6. Dear Robert,
    1.) Thanks for your articles on mutual intelligibility! Very interesting stuff; I think you shoud go for some serious and detailed research on the development of lingua francas such as Greater Turkic or Iberoamerican.
    2.) Occitan / Catalan in Google translate – this is probably due to orthographic convention. Google has a stupid robot brain, so it just doesn’t recognise words, aven if they are the same, just spelt differently. Why not go through two texts, one in Occitan, one in Catalan, to discover regularities in differences between consonants and vowels? This won’t cover everything, but it will get you a long way. It works well for German / Dutch – I am a German speaker, and you just have to always read “pf” for the Dutch “pp” and “ei” for the Dutch “ij”, and something like German starts to emerge.
    3.) Context – contextual clues are the ultimate tool for communicative competence, and even the most advance “semiotic” search functions of A.I. are nowhere near simulating the amazing abductive-intuitive guessing games that human cognition plays as a matter of course.
    If you say “I put you rsuitcase in the attic” with an upward-pointing hand gesture, or even just say it in a house where the attic is one of the most obvious stowing places, the addressee will be closer to guessing than otherwise.
    Keep the linguistics articles coming – fascinating stuff!

    1. Hi. Thanks Fergal. Did you see my articles on German where I broke German down into 168 different languages? If not, feel free to check it out and add any opinions you have on German dialects.

  7. Dear Robert, I don’t think that the automatic translators such as Google Translators are a good way to measure intercomprehensibility between languages. I as a French native speaker with a good knowledge of Latin am able to get the point of many (simple) texts written in another Romance language – except Romanian. But a computer is unable to do what a human brain is capable of doing: look for similarities beyond different orthographic conventions, don’t take details into account but rather the 70% of common roots and grammatical structures. The machine can’t make this kind of differences and would give too much importance to one misleading word. Just think of the problem of the two orthographic norms of Occitan – classical and Mistralian!
    on that topic, just a quote from Descartes’ Discours de la méthode : “[…] car on peut bien concevoir qu’une machine soit tellement faite qu’elle profère des paroles, et même qu’elle en profère quelques unes à propos des actions corporelles qui causeront quelque changement en ses organes, comme, si on la touche en quelque endroit, qu’elle demande ce qu’on lui veut dire ; si en un autre, qu’elle crie qu’on lui fait mal, et choses semblables ; mais non pas qu’elle les arrange diversement pour répondre au sens de tout ce qui se dira en sa présence, ainsi que les hommes les plus hébétés peuvent faire.”

  8. I have to say that I don’t think cross reading in Romance languages is that exaggerated. I began learning French using my Spanish and I found it to be incredibly helpful and made it easy because I started with reading subtitles on movies. I now speak French quite well. I think lexical similarity is where you realize how close the languages actually are. I have studied Italian, Spanish, French, and Portuguese and I can say that it does take a bit of knowledge to be able to read the other languages, but it is not impossible. i don’t think that i would be able to understand the written languages very well if I didn’t study them individually to some degree, but once you have done that just a little bit the rest comes with little difficulty. SP and PT don’t require much study, if you can read one, you can make out quite a bit of the other. I would note, however, that lexical intelligibility is very different than intelligibility of the spoken languages. All four of the most spoken Romance languages have a pretty distinct sound. I think that Italian and Spanish have the most similar spoken sound, but that has more to do with the alphabetic pronunciation than language/grammatical similarity.

Leave a Reply

Your email address will not be published. Required fields are marked *


Enjoy this blog? Please spread the word :)