Is this unique to certain families of language or all verbal human language?
The question can be interpreted as asking a lot of different things, and it's not clear what exactly you mean by "universal basis for consonants vs vowels". There a a number of schemes for describing speech sounds which purport to apply to all languages: basically, a number of feature theories, and the articulatory labels of IPA letters. IPA isn't a scientific theory, it's a convention, so it doesn't really make claims the way a feature theory would. The terminology used to describe vowels in the IPA is largely distinct from the terminology used for consonants. In that sense, IPA puts consonant and vowels on opposite sides of the fence, if what's what you mean by "versus".
In the realm of feature theories, the systems of Jakobson & Halle 1956 and Chomsky and Halle 1968 use a system of features that can apply to both consonants and vowels, although given the phonetic definitions of some of the features, certain values can't be assigned to vowels (for example neither [-continuant] nor [+anterior] can be assigned to a vowel). There have been many adjustments to feature theories, some modifications making vowel features and consonant features be more different while other make them be more similar. For example, for a while there was a distinction between "dorsal" and "velar", the former being about tongue position in vowels and the latter being about velar consonants, which in Chomsky & Halle are governed by [back] and [hi].
Another interpretation of the question would be, "can consonants and vowels always be distinguished in all languages", and the answer there is "Probably not". The Chomsky & Halle system does not have a feature that corresponds neatly to "consonant" and "vowel". It has a feature "consonantal", but that is different from "consonant" since [j w ʔ h] are consonants but not [+consonantal]. There is a feature [syllabic], where all traditional vowels are [+syllabic], but there are also syllabic sonorants like [r̩ m̩] etc. which are [+syllabic], yet consonants in some sense. Moreover, there isn't always some grammatical fact that puts [p d m l w s h] on one side of the line and [i ɪ a o] on the other. If you can decide what things you want to call "consonants" (glide being the problem) and what you want to call "vowels" (syllabic sonorant being the problem), then you can usually construct a description of those two sets in your theory.
Another possibility is "is there a functional explanation for having consonants versus vowels?", given that languages don't tend to just pile up string of consonants, or string of vowels, to make words -- there is a tendency to alternate consonant-vowel-consonant-vowel. There is a subtle force that pushes things in the direction of alternating CV, which is that CVCVCV enhances ability to perceive individual segments. The problem with consonants is that they tend to have little or no acoustic energy so they are hard to distinguish. The problem with vowels is that your mouth is open and you're leaking air rapidly, and moving from one vowel articulation to another is very slow and smeary, so it's hard to keep track of which vowels were uttered in what order in something like [ieɔuaʌɛæi]. If you interleave consonants and vowels in a pattern like [piketɔlupasʌtɛkæni], there is a very rapid and easy-to perceive change from consonant to vowel, and you slow down the leakage of air so you can talk longer. (That's the really abbreviated version of the functional story).