Old Dwarven language

The Old Dwarven language was the earliest attested variant of the Dwarven language, and it is the second most attested ancient language of Artograch that was spoken before the Proto-Elven invasion of Artograch. It is still used as for religious purpose in the Dwarven religion, although with modernized pronunciation.

Grammar
In the Old Dwarven language, words consisted of two main parts: roots and templates. Roots consisted only of consonants, while templates consisted of vowels and "empty spaces" - empty spaces that were filled by the consonants. For example, putting the root (metal) and the template  (people, folk), we get  (Dwarves, literally "People of Metal" or "People of Steel"). An example with the same template but a different root is (Elves), which uses the same template as before, but the root  (meaningless on its own, borrowed from the Proto-Elven "ethel").

Consonants
In Old Dwarven, each individual consonant came with three variants: Light (palatalized), Neutral (unmdoified/plain) and Dark (velarized/pharyngealized) - the velar series is a special case, in which the Dark consonants are uvular, while Light and Neutral are velar. The only one exception to this rule may be /w/, but that is special and may be also considered a vowel. The nasals and  appear as allophones of  before velar and uvular consonants respectively.

Old Dwarven had 36 consonant phonemes in total.


 * The voiceless uvular was pronounced as a fricative  in Steelhelm.

Vowels
In Old Dwarven, there were only three base vowel phonemes - or six if you count length as well -. However, their quality depended heavily on the surrounding consonants, producing 18 possible vowel sounds in total. Vowels surrounded by "Light" (palatalized) consonants have an advanced tongue root and are raised, vowels surrounded by "Neutral" (plain/unmodified) have a retracted tongue root and are slightly lowered, vowels surrounded by "Dark" (velarized/pharyngealized) consonants are pharyngealized and lowered heavily.

Sometimes, /w/ is considered a vowel too, because it's often part of templates rather than roots, allowing to appear after all three kinds of consonants.

This system produced a total of 18 vowels that can be seen in the chart below:

Pronouns
These pronouons are made up of roots and templates, like all Old Dwarven words:
 * The biconsonantal roots ṗ-q (meaning: self, ego), ť-r (meaning: opposite), ṡ-r (meaning: distance, distant)
 * The templates _a_á (meaning: person, individual) and _wi_i (meaning: people, group)
 * The end result is these six combinations:
 * ṗ-q + _a_á = ṗaqá (I, me)
 * ṗ-q + _wi_i = ṕwiqi (we, us)
 * ť-r + _a_á = ťará (you - singular)
 * ť-r'+ _wi_i = ťwiri (you - plural)
 * ṡ-r + _a_á = ṡará (he, she, it)
 * ṡ-r'+ _wi_i = ṡwiri (they, them)

Nouns
In Old Dwarven, every template was either singular or plural, meaning that there was a separate plural and singular version of each word. For example, the Old Dwarven word for a singular "dwarf" was (ḋaṙák), while the plurar "dwarves" was  (ḋwiṙik).

The explanation is that both words use the root ḋ-ṙ-k (literally meaning steel or metal), but the singular one uses the template _a_á_ (meaning person, individual), while the plural one uses the template _wi_i_ (meaning people, group).

In other words, the Dwarven word for a singular dwarf, "ḋaṙák" literally means "Man of Steel", while the word for plural dwarves, "ḋwiṙik" literally means "People of Steel".

This essentially means that anyone learning the language had to learn the singular and plural forms separately, as there was no consistent rule to relate plural and singular templates.

In Old Dwarven, there were no grammatical forms of words - instead, there were separate prepositions and postpositions to mark ownership, location, status as object or target of a verb, etc.