How do you hear vowels of other languages?

[krtk]

Although I grew up in an English speaking country, Finnish was my first language learnt in childhood, so I hear vowels in a Finnish way rather than an English way in spite of speaking English fluently. So for example it's very noticeable to my ears when foreign languages violate the Finnish rules of vowel harmony (to my ears it's particularly striking when listening to the Korean language, but even in languages such as Spanish where the /a/ vowel ranges between Finnish ä and a).

I'm curious whether others are also affected by vowel rules of their native language when listening to foreign languages. For example do native English speakers find it odd to hear /æ/ in open syllables in Finnish?

[Iversen]

Actually I long thought Finnish ä was pronounced as Danish æ or Swedish/German ä (i.e. close to the vowel in English "let"), but then I noticed that it actually is closer to a simple flat a. So in this this case I made an erroneous assumption based on the languages I already knew, but that didn't hinder me from eventually hearing that my assumption was wrong.

One example more: in a weird language called "English" 'y' is treated more like a consonant than a vowel, but you just have to go to Scots the see it used as a vowel (as in "Auld like syne") - but not as the same vowel as in Danish. Our 'y' sounds like German 'ü¨or French 'u'. And I do think I'm quite good at hearing the actual sounds without letting my knowledge of the phonemes interfere.

When I studied French I discovered (or was presented with, I don't remember) the general depiction of the possible vowels (shown below), and I have used that ever since to locate the sounds I heard in different languages. You can see it in the Wikipedia article, but as show there it doesn't indicate that it's doubled up by its nasal counterpart - and that there are things like halfvowels in for instance the Slavic languages as in the Croatian island name "Krk". But even those halfvowels can be placed in an image like the one in the Wikipedia article.

In some cases I have tried to compare the sound emitted by different native speakers of a language and more often than not found major variations. Like for instance in the Dutch diphtongs, which definitely didn't match the normative prescriptions in my books. The theoretically correct solution to that problem would be to choose one role model and adopt his/her pronunciation habits. But my solution to this problem has always been to develop my own mix of the things I hear and then modify that mix when I visit an area where the language is spoken. Perfectionists may scoff at this, but I don't mind speaking an idiolect based on compromises as long as native speakers understand the result.

Vowels.jpg

PS: I never cared to learn the IPA signs - I use my own homemade language dependent systems instead.

[Severine]

The only consistency I have noticed is that when I begin studying a language that makes use of sounds not present in my native language, I go through the standard phase of having difficulty identifying/distinguishing those sounds until I've had sufficient audio input to train my ears and brain to treat them as distinct and meaningful sound units. For example, the Russian Ы was my nemesis for quite a while. Hardly an original observation, but there you have it.

Although I'm not prepared to defend it rigorously as a method, I personally find it useful to engage in audio immersion irrespective of comprehension, i.e., spending time around native speakers or listening to talk radio even when I don't understand, just to let my brain soak up the sound library of a language and get used to its rhythm and music. It helps me build an idea of what I'm working toward.

[golyplot]

The only consistency I have noticed is that when I begin studying a language that makes use of sounds not present in my native language, I go through the standard phase of having difficulty identifying/distinguishing those sounds until I've had sufficient audio input to train my ears and brain to treat them as distinct and meaningful sound units. For example, the Russian Ы was my nemesis for quite a while. Hardly an original observation, but there you have it.

At least in my own case, it feels like that never goes away. I suspect that even on the occasions when it seems like I managed to tell a difference, it's really just my brain guessing based on context (i.e. recognizing the words and inferring the sounds based on that).

[orlandohill]

The only consistency I have noticed is that when I begin studying a language that makes use of sounds not present in my native language, I go through the standard phase of having difficulty identifying/distinguishing those sounds until I've had sufficient audio input to train my ears and brain to treat them as distinct and meaningful sound units.

At least in my own case, it feels like that never goes away.

Have you ever done any dedicated ear training? For example, the minimal pairs training found in the Fluent Forever app, or testing yourself on IPA recordings in an SRS?

[Iversen]

As I have written above I have spent some time listening for the actual sounds (phones) rather than the phonemes, which actually are theoretical constructs based on distinctive pairs - which means that each phoneme covers a selection of phones. But since I do my input in the form of written text I have also covered all the distinctive pairs I know of. So my answer to Orlandohill's questions is yes - except that I didn't use IPA or SRS.

PS: I have just listened to a video that claims that the soft d of Danish is so weird that no other language has got it . Well, maybe, maybe not. Anyway - I could live with the English th- from "with" in Danish words spoken by foreigners so the diference isn't that big in my opinion (but some people are just so fuzzy ). By the way, we have officially got 26 vowels (even without having any nasal vowels), but it could be worse: the language "Taa" (or "Xóõ") from Southern Africa has got 31 (according to some sources, Wikipedia says 5) - against just 13 in Standard English. At the other end of the scale, the Northern Caucasian language "Ubykh" is said only to have two vowels, but around 80 consonants (presumable including some halfconsonants that can be pronounced without the support of vowels). However the reason that it can make do with so few seems to be that those two vowels phonemes have divided the turf among them so that each covers a whole gamut of allophones. If that's true it shows in a nutshell why language learners can't just focus on learning to distinguish the phonemes.

[orlandohill]

As I have written above I have spent some time listening for the actual sounds (phones) rather than the phonemes, which actually are theoretical constructs based on distinctive pairs - which means that each phoneme covers a selection of phones. But since I do my input in the form of written text I have also covered all the distinctive pairs I know off. So my answer to Orlandohill's questions is yes - except that I didn't use IPA or SRS.

Listening for phones rather than phonemes sounds like a very interesting approach, especially given the wide range of pronunciation differences between dialects of some languages. The use of multiple speakers would make it a form of High Variability Phonetic Training.