Here are mine:
- The consonant inventory often ends up using one of these two as a basis:
- One inspired by ancient Greek, with tenuis, aspirated, voiced, and nasal obstruents at three places of articulation. A major difference from it that I seem to like making is that, instead of just the one /s/ for the fricatives, sibilants behave just like plosives and can also be tenuis, aspirated, voiced, and nasal (though /z̃/ is given a habit of not wanting to sticking around).
- The plosives are like SAE but they have corresponding glottalic versions as well, maybe missing otherwise expected /pʼ ɠ/.
- /a e i o u/ are nearly invariably present, at least at some point. To this I may add /ɛ ɔ/ or /y ɯ/ or, if I don't want them in the earlier stages yet, delete /e o/. Length distinctions may or may not be present, but they often are for smaller vowel inventories. When I do have length distinctions, a conversion from that to a tense-lax distinction is on the common side.
- Where /θ ð/ show up, they're remarkably stable if Let's Reform English is any indication. (What? I like those sounds )
- An avoidance of more complex syllable structures than CCVC.
- The phonotactics rules tend to be very simple on a morphophonemic level, often behaving on the syllable and not the whole word (though sound change may have something to say about that). Onset rules are fairly liberal (syllable-initial /ŋ/? Why not?), with plosive+approximant clusters being par for the course, though I will sometimes have a lang force there to be an onset for every syllable. My three most common rules for codas are any consonant, only certain types of consonants (I usually permit at least /p t k m n ŋ s/), or none at all.
- Nearly every language has a case system, with nominative-accusative and active-stative my two favorites to use. As regards treatment of ditransitive verbs, secundative languages show up noticeably more often than in real life, but they're not the majority. Suffixaufnahme is not unknown for genitive constructions.
- Nouns may have gender, but it's less common than in SAE. When they do, though, the neuter is more common than in SAE. My older conlangs that had gender occasionally also had a rule where third-person genitive/possessive pronouns had to agree with BOTH the subject AND the object (i.e. my words for "his", "her", and "its" all had masculine, feminine, and neuter forms).
- Higher numbers tend to be formed like in East Asian languages rather than there being separate words for multiples of powers of ten and the first few numbers past ten. (Replace "ten" with "twenty" if I decide I want that base instead, in which case expect there to be a sub-base of 5.)
- Fusional languages invariably trace their origins from an earlier agglutinative stage.
- Typically no infinitive; the base form of a verb is the third-person singular.
- Causative constructions are more productive than in SAE, especially when deriving verbs.