The database includes four levels of borrowing certainty, based at times on rather tentative judgements.

(i) Very little evidence for borrowing

applies mainly to so-called “areal roots” which we consider to be of such old age in the languages concerned that they can be reconstructed in more then one language phylum, either due to early borrowing (with unclear direction in terms of donor and recipient languages), or due to universal onomatopoetic properties of the word form itself. An example of the latter is tóofàa ‘to spit’, for which many similar equivalents in form and meaning can be found across various language families and phyla without, however, being able to identify any direction of borrowing, if at all (it could be accidental due to its onomatopoetic features). This judgment would also apply to lexical similarities in terms of form and meaning across rather large geographic distance, such as, for instance, with jínkáa ‘the thatching’ which looks a bit like the word jonge ‘the thatching’ in Mande (Niger-Congo) which is a less likely contact language.

(ii) Perhaps borrowed

applies to words which, on the one hand, would appear to be reconstructable to Proto-Afro-Asiatic (PAA) yet lacks regularity in terms of sound changes; cf. the noun sâa ‘the ox’ ín Hausa to be compared to *hla in PAA. However, Newman (1977) suggests that regular phonological change of PAA lateral fricative *hl should give initial l or r for Hausa (cf. the etymological reflex in closely related Angas: lon). This judgement also applies to words which other authors have considered likely loans, somewhat contrary to our own opinion, such as with the verb làasáa ‘to lick’ for which Jungraithmayr / Ibriszimow (1994) maintain that “ungraded Hausa 'lààsáá' may be of Arabic origin, since initial 'l-' in Hausa does not point to great antiquity”. We have also applied this judgment to words which strongly resemble words in probable contact languages, even though their onomatopoetic features could point into the direction of rather universal distribution (cf. “very little evidence for borrowing” above):cf. Hausa àgwàagwá ‘the duck’ and compare it with gbangban, gbamgba ‘the duck’ in Nupe, both could be onomatopoeic in origin and need not represent borrowing. In other cases, borrowing is a likely hypothesis, if we are not dealing with chance resemblance, cf. Hausa ícèe ‘wood/tree’ and idi ‘tree’ in Tubu (Saharan).

(iii) Probably borrowed

applies to some probable Arabic loans which, however, don’t violate assumed phonotactic rules of Hausa (and Chadic languages in general) and, therefore, could be considered common heritage from Proto-Afroasiatic but were we have judged them to be Afroasiatic-internal loans instead. This judgement is also applied to words which again don’t necessarily violate phonological patterns of Hausa but which are also present in some contact languages where they appear to match the phonotactics of the languages even better. We have further applied this judgement to words which are present in likely contact languages but where we are not sure about the direction of the borrowing.

(iv) Clearly borrowed

applies our highest degree of confidence with regard to words which we think we know or can very plausibly assume to be borrowed into Hausa from a donor language even though the ultimate donor or the intermediate languages, if there were any, cannot be identified. Insofar, even this rather strong judgment remains rather tentative in the light of the rather poor quality and quantity of available data and the almost complete absence of systematic comparative historical linguistic work in the wider area in which Hausa is spoken.

The database contains the closest equivalents of LWT meanings in Hausa also in terms of general usage as judged by the near-to-native speaker competence of one of the investigators (Ari Awagana). Dictionaries such as Bargery (1934), Newman (1990) and Awde (1996), have been consulted for specific words, i.e., usually for words which occur in rather formal speech and are rarely used but which are assumed to be understood by educated speakers. This is particularly the case with some loans from Arabic which, however, can be assumed to be known by Hausa speakers who have been through the widespread early-childhood Quranic education. Also, we have chosen to incorporate “modern” terms that are typically used in the media based on the assumption that these terms are known to practically all Hausa speakers. With regard to words that can be considered rather local or regional (“dialectal”) and for which a more widespread term is available, these have been avoided and not been incorporated into the database.

As regards transcription of Hausa words, we are not following the official orthography which marks neither tone nor vowel length. We have adopted conventions used in Hausa linguistics which mark tone by diacritics on top of vowel symbols (á = H[igh] tone, à = L[ow] tone, â = HL Falling tone) and which mark long vowels by doubling the vowel symbol (aa = long a) rather than using a macron. Note that we, therefore, disallow tonally unmarked syllables which, however, would reflect a widely used convention for indicating a High tone syllable in Hausa. Note further that with long vowels, the diacritic for tone is only used once and on the first of the two vowel symbols (áa H tone, àa L tone). According to the conventions used here, we, therefore, write the word for “world” in Hausa as dúuníyàa rather than as dūniyā̀ (as other conventions would allow).

A meaning is entered in this field only if there is a significant difference between the LWT meaning and the Hausa equivalent, or to explain shades of meaning in case of multiple equivalents of LWT meaning.

The following criteria have been applied in terms of describing the analyzability of a word form in Hausa.

(i) Unanalyzable:

Applies to simple monomorphemic words.

(ii) Semi-analyzable:

Applies to cases in which at least one of the analyzable parts of the word form has taken on a specialized (grammaticalized) functional meaning, or the word form does not correspond to the formation pattern(s) found elsewhere in the grammar. This can be illustrated by the following examples:

(a) complex numbers like goomà shâa bíyú ‘12’, góomà shâa úkù ‘13 ’, etc. in which the part shâa (probably etymologically identical to the polysemous verb ‘to drink’) functions as a connective particle;

(b) highly lexicalized noun plural forms like mútàanée (people) from mùtûm (human being), ítàacée (trees) from ícèe (piece of wood) etc. which allow infixation (such as -aa-) and are accompanied by considerable morphophonemic changes, such as m <> n (as in mùtûm > mùtûn > mút-àa-n-ée) and t <> c (as in *ítèe >ícèe, and *ít-àa-t-ée > ít-àa-c-ee);

(c) CVC-reduplicated forms like ɗánɗànáa (to taste); mùr̃múshíi < *mùs-mús-íi (smiling);

(d) juxtapositions which do not conform to the usual “genitive type” patterns for noun compounds in Hausa like ùbáa kíntàa (the stepfather) which lacks an expected “genitive linker”, cf. also ráanáa tsákàa ‘midday’ (lit. ‘the sun + the mid’);

(e) verbal derivatives with linguistically still disputed morphological structure, such as the so-called Grade V “efferential/causative” marking sáyár̃ dà ‘to sell’ from sàyáa ‘to buy’, in which the morphological status of the apparent “preposition” dà remains somewhat doubtful in the light of evidence from within Hausa grammar.

(iii) Analyzable derived:

Applies to fully regular Hausa derivations such as agentive nouns (máhàifáa ‘parents’ from hàifáa ‘to give birth’), instrumental nouns (mátáuníi ‘molar tooth’ from táunàa ‘to chew’), locative nouns (másálláacíi ‘mosque’ from sállàatáa ‘to pray’), abstract nouns (mâitáa ‘the magic’ from máayèe ‘the sorcerer’), etc.

(iv) Analyzable compound:

Applies to compounding by regular genitive construction involving a gender-sensitive linking morpheme, such as in bàakí-n ruwaa ‘shore’ (lit. ‘the mouth/fringe-of water’, with linking morpheme -n), or in some kinds of adverbial constructions which also make use of a genitive-type construction, such as ná ƙàrshée ‘last’ (lit. ‘of the end’, with linker ná), cf. also cardinal numbers that use the same construction: ná úkù ‘third’ (lit. ‘of three’). This also applies to certain constructions which make use of the derivative elements mài ‘the one who own/masters/makes…” and its polarity counterpart máràs ‘the one who lacks…’ (both ultimately from agentive constructions involving the identical prefix ma-), cf.

mài túkwàanée [WHO OWNS/MAKES pots] ‘the potter’

mài hánkàlíi [WHO HAS cleverness] ‘the clever one’

máràs náuyíi [WHO LACKS weight] ‘the one light in weight’

(v) Analyzable phrasal:

Applies, for instance, to many items in Hausa which involve verb-noun constructions, such as yí bárcíi ‘to sleep’ which is ‘to do sleeping’ (with bárcíi as a noun and yí ‘to do’ as a more or less dummy verb), Other verb+noun constructions use rather specific verbs, e.g. yánkè húnkúncìi ‘to adjudicate’ which is ‘to-cut-off sentence’. Hausa has a lot of highly descriptive verbal compounds which resemble whole clauses, such as máalàm bùuɗè líttáafíi ‘the butterfly’ which is ‘teacher open(s) the book’. Prepositional phrases often translate into single lexical items in English, such as à kàryé [PREP break.PRESENT.PARTICIPLE] ‘broken’, dà wúríi [with place] ‘early’, dà cíkìi [with belly] ‘pregnant’.

In terms of our rather tentative guesses on the age of loans, we use the following chronological approximations:

Proto Afro-Asiatic period: more than 10,000 BP

Proto Chadic period: 10,000 – 5,000 BP

Proto West-Chadic period: 5,000 BP

Ancient Areal Roots = Proto-Chadic period

For the Hausa language, we use the following rough approximations to relevant periods:

Pre-modern Islamic period 1000 - 1900

Pre-modern non-Islamic period before 1900

Pre-modern Kanuri contact period 1300 - 1900

Modern (colonial & postcolonial) period 1900 - date