Gender and noun class
In typology, grammatical gender is a classification of nouns reflected in agreement. It is not identical to biological sex and need not be marked on the noun itself. WALS chapter 30 counts genders by the agreement patterns nouns control, and explains why counting singular and plural noun-class forms separately can give a misleadingly larger total. To establish a system, compare agreement targets under controlled conditions, not a list of noun endings.
A short corpus may contain only one agreement construction or mostly proper names. In that situation, an apparent two-class distinction could be gender, number, declension class or lexical coincidence. State the distribution and competing analyses. If the language lacks enough attestations, the value is unknown rather than “no gender.”
Definiteness
English the is a separate definite article, but that is one strategy among several. WALS chapter 37 counts distinct words, article-like demonstratives and noun affixes under its definition; some languages have no dedicated definite article. The comparison asks how reference is signaled in the construction, not whether speakers can distinguish known from new entities. A definite interpretation in translation does not prove an unwritten article in the source.
Demonstratives may take on article-like use over time. For a historical sequence, compare dated examples and discourse conditions. Do not treat a later grammaticalized article as proof that every earlier demonstrative occurrence already had that function.
Evidence source
Evidential markers signal how information is obtained: directly witnessed, inferred or reported are broad distinctions. WALS chapter 77 counts grammaticalized evidentials and explicitly excludes ordinary lexical adverbs such as reportedly from that map. It also separates evidence source from a speaker’s degree of certainty. A confident report can be indirect; an uncertain visual observation can be direct. A translation like “apparently” cannot identify the morphology on its own.
Exercise: choose one article candidate and one reported-information marker from a grammar. For each, write the distributional test that the WALS chapter requires. If the published evidence does not meet the test, record the gap rather than forcing a map value.
Modern-language examples on this page clarify a linguistic pattern. They do not establish unattested sounds or forms for an ancient language. Follow the cited sources below and keep editorial readings separate from the surviving trace.