There are two types of 'neutrality' that we can discuss in relation to the IAL argument; one is linguistic neutrality, which seems to me to be one major aim of Sona (at least the book's subtitle seems to say so), and cultural neutrality, which is a principle aim of Lojban and a partial result of its being based in logic. One problem is, it seems like there is no sound definition of what it means to be linguistically 'neutral'. In Zamenhof's time and place, a language with roots plucked at random from different European language groups, with only the most basic European style grammar and free word order, is fairly sufficient. Searight, living in a slightly more global time, largely tried to accomodate the features of Asian and possibly Middle Eastern languages as well as some features of the European ones. But are either of these cases of true neutrality, or just compromise? One could try to argue that Lojban, for example, can be regarded as more neutral than either Esperanto or Sona because its grammar is largely based on predicate logic, and logic is an inherent capacity of all humans (indeed, Esperanto in its time was seen as a more 'logical' form than any existing European language); but of course Lojban's grammatical structure does not derive purely from logic, and even if it were not influenced heavily by English, may arguably be influenced by computer programming languages (?!) or from the Western perspective on linguistic analysis. Cultural neutrality, on the other hand, may be more obtainable. I think lojban does as good a job any any language in reducing cultural artefacts in language; there are certainly no elementary expressions carried over from idioms of existing languages (Esperanto cannot say the same for itself), and the ways in which meanings can be evaluated is fairly regularly defined, restricting unwarranted metaphor in most places. Having laid out these issues, I think the most fruitful project would probably involve something with a very formaly constructed grammar like Lojban, but at the same time taking into consideration the importance of these central problems: 1) Being able to teach and learn the language in a short amount of time; fewer root words and high regularity means less memorization. (related to the 'ease of use' argument) 2) Being able to unambiguously and, in the general case, with minimal effort determine the meaning of an utterance on the basis of clearly defined grammar rules. 3) Within the limits of (1) and (2), still being able to express the wide range of concepts that are prone to be the subject of international communication (which, in effect, ought to be any concept at all). Sona clearly is strong in (1) but very weak in (2). Lojban is very strong in (2), and is fairly competitive in (1), but I wager that it can still be improved; Lojban does only have a few thousand root words, but even that can be a lot for some people (including me) who will determine whether or not they choose to learn a language based on how long it will take to get good at it. Also, Lojban's system of word compounding is not so well defined (one of the aspects in which Lojban is unusually weak at (2)) and, frustratingly, roots can often change form if they are compounded (I'm pretty certain we don't want a widely-accepted IAL to do this).