The audio makes a lot of mistakes but it's actually correct here.
Most of the time, ة , which is only used to mark feminine nouns and adjectives, is just pronounced as "a", but classical/standard Arabic has and extra u/a/i at the end of nouns and the adjectives that describe them to mark subject/object/relation-to-other-nouns respectively, and when ة is followed by one of those vowels, it is pronounced "at". Then the "n" you hear at the end marks that she is AN engineer, not THE engineer.
Because Arwa is a proper noun, it never has any of these extra sounds at the end.
So overall we have Arwa muhandis+at+a+n = Arwa engineer + femenine ending with t pronounced because of the following vowel + object marker + indefinite marker
However, as opposed to formal Arabic (AKA: fusHa) in vernacular dialects, the case marking vowels and the n that marks indefiniteness aren't pronounced, so neither is the t in this case, and you just get muhandisa.
The reason the computer gets it wrong so often (bit not this time) and seems so irregular is because it's switching between very formal and informal within a phrase and it really shouldn't. Nouns and adjectives must have the same endings and agree with each other. So all of them must have those extra vowels and "n"s or none of them should.
Note that although vernacular dialects do not use case marking vowels or the indefinite n much anymore, the ة still gets pronounced as "at" rather than "a" when feminine nouns combine with other nouns to make compound nouns.
Oh yes. Of course. Thanks.
I'm a second year student, and all of the computer's mistakes in noun/adjective agreement are really testing whether I can tell a grammatical sentence from an ungrammatical one.
I imagine it would be really hellish to try to figure out how to say things correctly by listening to the audio for a complete beginner.
yes, it would. Had this long discussion with one of the other course contributors on some threads (you will eventually find it :) ) who was kind to engage, and defended the choice of the erratic grammar as 'finding a middle ground between MSA and dialect' when I really disagree because there is no such thing (and there is no single 'dialect' either) neither is that how TV presenters in international Arabic stations speak (like BBC Arabic) - it is always MSA.
The course needs a paradigm shift to adopt proper MSA, or to pick one specific dialect and teach it.
I have seen a strong bias of this course towards Levantine dialect (a beautiful language in its own right, but not 'Arabic'), spoken by less than 10% of Arabs. In vocabulary, structure, and pronunciation. It is, in my view, unfair to both the learner and to the language.
For some reason I can't respond to the lower comments, so I will respond here. Thank you for your explanations FiX and Bryson610892. I'm a beginner. I was under the impression that Duo created a tool for MSA/fusHa, but hearing that the contributors couldn't keep their own dialects out of the material really turns me off to using Duo. I need to learn MSA/fusHa, not a jumbled mess. There's no "middle ground," just bad teaching.
I also appreciate the clarification of FiX et al., but am sticking with Duo's beta Arabic for now because I don't even know the alphabet, so any assistance in getting started is a plus for me. I'm then supplementing with grammars with the hope that in a year or two Duo will do another iteration of this course that takes into consideration the fine points that are being brought up.