Blind society: language basics

I've been getting frustrated with the fact that I can't name my blind society or give any of their words for anything, because I don't have any of the language invented. So I started coming up with some of the basics for it this morning. All this is very much first-draft and subject to revision.

Morphologically, it's largely isolating with regard to inflection, but it does have derivation and compounding. Kind of like Ikanirae Seru (my best-developed conlang) that way, and not entirely accidentally - it's what I find easiest, even if it's not what I like best. (Over the course of time, some independent words may turn into inflectional affixes, but that hasn't happened in the main period I'm focusing on.)

I don't know anything about syntax yet.

For phonology I'm drawing on Pacific Northwest native languages, Proto-Indo-European, and Semitic, but not relying only on them for possibilities. I don't have the full phoneme inventory worked out, but I have a lot of the ingredients.

The vowel system is a basic 5-vowel system.
Syllable structure is C(C)V(C): Onsets are obligatory, initial two-consonant clusters are possible, and single-consonant codas are possible.

There's a glottalized/unglottalized division running through pretty much the whole consonant system, including sonorants as well as obstruents. (Glottalized obstruents are ejective, but it's manifested differently in sonorants.) Obstruents additionally have a voiced/voiceless distinction in unglottalized sounds.

Places of articulation for obstruents and nasals include labial (probably bilabial for stops and labiodental for fricatives), dental/alveolar, velar, and labiovelar. Fricatives may make more distinctions in the coronal area (i.e., in addition to /s/ there may be /θ/ and/or /ʃ/, and their voiced and glottalized counterparts); I do know that there's an alveolar lateral fricative /ɬ/. I'm not sure if there are affricates, or if affricate-like clusters just come together when a fricative follows a stop.

For non-nasal sonorants there are /l/, /r/ (not sure what kind of r yet), /w/, /j/.
There's also glottal stop and /h/.

There are phonotactic restrictions and phonological rules relating to how different laryngeal specifications combine in clusters. They can almost be summed up by saying that there can only be one distinctive laryngeal specification in a cluster, though that's probably not quite true when sonorants are involved.

If any consonant in the cluster (sonorant or obstruent) is glottalized, the whole cluster counts as glottalized; although the glottalization may be more manifest in one part of the cluster than others, there are no distinctively non-glottalized sounds in the cluster.

When there are no glottalized consonants involved, if any obstruent in the cluster is voiced, all other obstruents in the cluster are voiced. Clusters with multiple voiceless obstruents only occur when both obstruents are underlyingly voiceless.

Since sonorants don't have a voicing distinction, a voiced (non-glottalized) sonorant can combine with either a voiceless obstruent or a voiced obstruent in a cluster.

I'm not quite sure how /h/ and /ʔ/ fit in with these rules.