LyzTyphone:
So you mean devoicing doesn't normally occur in Esperanto? Well, that's a bit disappointing, as I really want to have an example with Esperanto in our Linguistic class to give my classmates (and teacher) a bit of a taste for the language. I think if they have seen the Esperanto they might want to look deeper into it.

It's not supposed to happen. I think you could generate some interest, however, simply by discussing how Esperanto, as an L2 without a normative L1 speaker base, resists tendencies, such as devoicing, found in other languages.

Allophonic variation with position in the word may occur in Esperanto. But in your examples Liz [ekiri] [egbatali], you crossed phonemic boundaries.

I think that the approach suggested by Todd is a good one and should interest your teacher.

LyzTyphone:You do realize that strict phonologically speaking, the "k" in "ka" "ki" "ku" are actually different sounds, right? This is (yet again!) quoted from Pleana Analiza Gramatiko, that I just learned yesterday. There is no perfect phonological language in the world. Esperanto is almost there, as it doesn't have a lot of harmonizaion or phonological alternation seen in other languages, but it's still not one. And it's perfectly Ok as long as the phonological deviations doesn't cause difference in meaning (remains allophones of the same phoneme, by the jargon).

(Not@Lyz, rather for anyone coming by and going "THE EO PROPAGANDA MACHINE LIED TO ME EO'S NOT PHONETIC!)
EO's still a perfectly phonemic language since there is no reason to differentiate between the k in all 5 flavours (see here).

Actually, for the scientific minded, it's probably best that we berate ourselves into using just "phonemic" to describe EO, since phonetic is a misnomer and would give learners hernias over how they have to speak like perfect machines?

@ceigered
actually this might have nothing to do with mandarin, as it lacks a differentiation between and but instead distinguish consonants by and [kh]. I think I actually use that when I speak Esperanto and, to a lesser degree, English by corresponding my with , and my [kh] with . And I haven't met people who say they can't understand my word because of that.

I meant that because of the way people from different backgrounds perceive different sounds, they might subliminally get away with some differences in their speech. E.g. a Mandarin speaker might be able to subconsciously pronounce ekbatali with their version of a "g" (in fact an unaspirated k), but if I used my version it would stick out like a sore thumb since I have a very unrefined accent where my "g"s are very loose sounds, not to mention the voicing is much stronger (or perhaps it's more that there's less aspiration or some sort of palatalisation, who knows; if my English were to be the standard, the language would be impossible for foreigners )

Thus a Mandarin speaker's [ekpathali] might be more understandable than my [egbathali], despite the fact that the speaker's (and all other unaspirated stops) normally corresponds to an EO /g/ (and other EO voiced stops).

That's what I was trying to get at (although I don't think I wrote it well, even then too)

EDIT: Perhaps a better comparison would be the Mandarin /g/ vs the Japanese /g/ (in some accents: [g~ng] - can't be stuffed finding the actual IPA symbol on this comp).

I shouldn't be too disappointed Liz. There is scope here for a very interesting psycho-linguistic experiment.

If you can earn extra points in your course with a project. You might suggest to your teacher that you create a sound file of Esperanto words (both genuine and ones not in the dictionary) containing consonant pairings relevant to the topic, and invite Esperantists from all over the world to write down what they hear and email the transciptions to you.

However this comes out, the results are of academic interest. Either it will prove that Esperanto doesn't have the assimiliations of other languages, or, and this would be particularly interesting, it will demonstrate that perception is conditioned by the meaningfulness of the words.

So you can imagine that perhaps 'egzameno' might be transcribed as 'ekzameno', but 'egzajmi' might be transcribed as 'egzajmi'.

It always amazes me that the academic world has done so little in the way of experiments on Esperantists. It's a wonderful testing ground for linguistic notions, and it is so easy with the Internet to find willing subjects for the experiments, with vastly different L1 backgrounds.