Are there common traits of foreign accents across languages?

Question

A foreign accent is obviously influenced by the native language of the speaker. What I'm wondering is if there's been any research on common traits of foreign accents across languages, that one could find even when not being familiar with the phonology of the language being spoken nor the native language of the speaker, something that is common to not speaking your own language. I'm thinking something like hesitations or phonological inconsistencies.

Edit: The reason for this question, I must admit, is that hearing people speak languages other than their native language really annoys me. I wish it didn't, and I'm wondering if there's anything to it other than me being judgmental. Hearing a person first speak a foreign language and then switch to their native language I am often struck by how the way they speak their native language is so much more natural and immediate, even when their vocabulary and pronunciation of the foreign language are at near-native level. This is all of course very impressionistic, so it'd like to read what's been written about it, except I don't know how or where to look for it.

Answer 1

I don't know of research, or whether the following is true, but phonological alternations might be due to phonological processes in the speech of native speakers, but due instead to phonological rules in the speech of nonnative speakers. The distinction between rule and process is made in David Stampe's theory Natural Phonology. Rules have to be learned, and it's an error not to apply them when a language has them; processes are due to the arrangement of articulatory organs in humans, so they do not have to be learned, and it's an error to apply them when a language forbids them.

Here is a hypothetical example. Many languages devoice word final obstruents (oral stops and fricatives), e.g., German. This is a natural process. English forbids this -- we have to be able to say word final b/d/g/z/v in English. Japanese does not forbid this devoicing -- there is no reason to, since Japanese has no word final obstruents. However, it's been observed that Japanese speakers learning English do often at first devoice English word final obstruents. This error of devoicing where you shouldn't cannot be something Japanese speakers have learned about Japanese -- it is not language interference.

Now, I don't know what adult English speakers do when they learn German, but it seems plausible that they learn to devoice word final obstruents as a rule, since they must have suppressed the process of word final devoicing when they first learned English (though some children learning English do apply this process, for a time).

So, then, I'm guessing that a person speaking German who occasionally forgets to devoice a word final obstruent can be identified as a nonnative speaker of German.

Answer 2

In general, no, it is not possible to identify a foreign accent in a spoken utterance when you do not know the phonology of the language spoken nor the native language of the speaker.

As Greg Lee indicated in his answer, there are certain natural tendencies in languages of the world which are likely to emerge under such circumstances (like the final devoicing) but you have no way of telling whether they are part of the language in question or not because since they are natural, they are pretty likely to be a part of its phonology in the first place.

The same applies to the phonological rules of the native language (as an example - Spanish speakers tend to expend initial /sC-/ cluster to /esC-/) - again you have no way of telling from which language this comes from if you lack the knowledge of either. This is also fairly logical since you have languages from the same family where such a rule did not evolve (e.g. Italian), then you have languages where it did evolve and is still active (Spanish and Portuguese) and then you have language where it did evolve but stopped getting applied (French).

That being said, there might be some traits pointing out to the fact that what you hear is a non-native speaker with an accent, e.g. I imagine that the overall number of sounds distinguished by the speaker would be, in average, lower than in the spoken language but again, this is difficult to establish when we speak about sounds and not phonemes (remember, we do not know phonology) and furthermore it would not be more than an indication of the possibility, not a reliable identifier. Another feature that could be used for the same purpose is the speech pace, which would roughly correlate with the level of language skill and thus inevitably at the lower level with accent.