I'm currently looking at an old past paper which asks why it would be impossible for a human vocal tract to produce a pharyngeal nasal and a uvular click. Could anyone help me out?
Reflect on the meaning of the terms. A click is produced by making a velar closure simultaneously with a closure in the front of the mouth, then creating ingressive air by retracting the tongue body. But uvulars are behind velars, so a uvular click is impossible.
I don't see a problem with a pharyngeal nasal, though, so long as you don't try to make a uvular click at the same time.
In the case of a "pharyngeal nasal", it depends on what you mean by "a nasal" and what you mean by "pharyngeal". In the case of [m] for example, there is a complete constriction at the lips and the velum is lowered [m] is a "nasal stop". There are also nasal non-stops. There is a lot of equivocation over what is "pharyngeal", since most "pharyngeals" are actually epiglottals. There do not appear to be any purely pharyngeal pure stops (for whatever reason), and if a "nasal" is necessarily a stop, there can't be a "pharyngeal stop". On the other hand, if you use "nasal" to include approximants and fricatives, and don't care strictly about pharyngeal vs. epiglottal, then the articulation can be made. I don't think we can compute whether such an articulation would have detectable acoustic consequences upon which a contrast could be founded, but there's a good chance that such a sound would be be acoustically distinct enough to be phonemic.
Clicks require a constriction somewhere at the back of the mouth, plus some further forward constriction. The back constriction could be anywhere between on the soft palate back to the uvular region. Some of the Khoisan languages (esp. !Xóõ, Jo !'hoansi and N|uu) have uvular clicks, the latter language having been the subject of ultrasound studies by Miller. In fact, the back constriction on Zulu clicks moves from velar to uvular, and there is mounting evidence suggesting that the back constriction is clicks is commonly uvular. See Miller "The representation of clicks".