Zero-derivation entries in a Setswana dictionary

Zero derivation in a Setswana dictionary concerns a form being used as a different word class without an overt derivational marker. It is narrower than ordinary grammatical variation: an inflected verb form is not a new lexeme, and an English loan used as a Setswana verb is not automatically a converted word. The editor must identify the pattern, describe it accurately and make the supporting evidence visible. Comparisons with English forms such as book, noun and verb, can clarify the idea, but they do not provide a ready-made analysis for a language organised by noun classes and different verbal constructions.

A dictionary is both a historical record and a working instrument for speakers. A poorly handled zero-derived entry can hide a productive pattern, split one word into misleading pieces or attach an invented suffix to a form that speakers simply use unchanged. Setswana has speakers and publishing traditions in Botswana and South Africa, as well as communities in Australia, so the record must accommodate regional variation, language contact and present-day usage without treating one variety as the only standard. English explanations written for Australian readers should use familiar Australian English while retaining the grammatical precision needed for specialist consultation.

What zero derivation means in Setswana

Zero derivation is a relation between lexical forms in which one can be assigned to more than one word class without an overt marker. It is not simply polysemy: related meanings can remain within one class, whereas zero derivation crosses a grammatical boundary. Nor is it ellipsis. An editor needs evidence that speakers recognise a systematic relation rather than relying on an isolated resemblance between a noun and a verb.

Setswana makes this distinction especially important because noun forms can be identified through class markers, agreement and syntactic position. Candidate pairs such as loma ‘dream’ and bua ‘work’ must therefore be tested against actual noun and verb uses rather than assumed from translation equivalents. A form followed by a nominal suffix is not zero-derived merely because its base can also occur as a verb; the suffix changes the analysis. Nor should a missing noun-class prefix be mistaken for zero derivation: zero class marking may be an allomorph of the noun system rather than a conversion from another class.

Three tests help organise the decision. First, distributional evidence asks whether the form takes noun-class agreement, determiners or plural marking. Second, syntactic evidence examines the complements, tense and mood it selects. Third, semantic evidence asks whether a nominal use names an event, artefact, person or experience. A form may pass one test and fail another, which is why a single example is unsafe. Where a style guide uses the label infinitive, a note on the phrase ad infinitum can orient readers, while the entry itself should describe Setswana morphology on its own terms.

Deciding whether forms belong together

The main structural choice is whether the two uses share an entry. A related noun and verb with a transparent, productive relationship can appear under one headword with separate parts of speech, as English walk does. Separate entries may be preferable when the senses have unrelated origins, incompatible syntax or distinct histories. A cross-reference can then connect them without pretending that shared spelling establishes derivation. The entry policy must be stable enough that readers can distinguish conversion from homonymy, a source variant or a noun formed with an overt suffix.

Etymological checking can prevent a false grouping. The account of mogolwane’s Southern Bantu etymology shows the value of comparing related forms across Southern Bantu before assigning a historical label. The same principle applies when older material contains an affix that is no longer productive. Historians may explain how a form developed, but the modern dictionary should not represent an absent segment as if it were still visible to the speaker.

For each candidate pair, the entry should state the relationship plainly: converted from the verb, used as a deverbal noun, or recorded as a separate homonym with no demonstrated link. That wording should remain consistent across print, web and linked datasets. Contact forms require particular care. An English-based compound, an Afrikaans calque and an inherited Setswana conversion can look alike in a translated example while having very different histories. Borrowing, calquing and zero derivation should never be collapsed under a generic label such as ‘modern usage’.

Building evidence from usage

A reliable decision starts with a corpus designed for the question. It should combine Botswana and South African Setswana across newspapers, books, formal documents, learner materials and natural conversations while recording place, date, register, speaker age and orthographic choices. Private online exchanges require consent, restricted storage and anonymisation; frequency alone is not a licence to expose a speaker. Oral sources also need consistent transcription and an explanation of whether spelling represents speech, a published norm or an editorial normalisation.

A compact survey of Setswana weather words illustrates why semantic fields are useful test cases. Words such as pula ‘rain’ and seetse ‘heat’ belong to one weather network, yet membership in that network does not determine grammatical class. A corpus can reveal whether a form is used independently as a noun and a verb, whether one use requires a demonstrative or noun-class marker and whether apparent examples are really quoted words or ordinary constructions. The lexical set supplies promising candidates; full sentences supply the analysis.

Every illustrative citation should preserve enough context to test the claim. Include the date, source and a concise grammatical gloss, then use ellipsis where the omitted words do not affect the analysis. A written example can be supplemented by a transcript reference without reproducing personal information. Australian materials, including AustLit records and material located through the National Library’s Trove service, can show how Setswana appears in local schooling, publishing, churches and migration settings. They should supplement rather than replace evidence from Setswana-speaking communities.

Working within an Australian lexicographic context

Australian English offers useful precedents for documenting conversion, but local usage must be identified as such. The Macquarie Dictionary, prepared in Sydney, and the Australian National Dictionary associated with the Australian National University in Canberra both show how Australian meanings and spellings differ from British and international models. A Setswana dictionary prepared for Australian users should likewise label regional Setswana usage plainly and avoid presenting contact-era forms in Britain, Botswana or South Africa as interchangeable.

Australian speech also reminds editors that frequency and informality do not determine lexical analysis. Forms such as arvo ‘afternoon’, barbie ‘barbecue’ and the pragmatic expression no worries ‘it is acceptable’ have been discussed as part of local usage. Their different histories—clipping, abbreviation and a routine phrase—show why a shared function cannot prove a shared process. Melbourne’s public libraries and community centres, where bilingual story sessions and language-exchange activities bring many language groups together, also create a practical setting in which community reviewers can test whether an explanatory note reflects ordinary interaction.

Translation requires a parallel discipline. The analysis of Bashō’s translated haiku shows how an English rendering can reproduce an image while changing its grammatical force or cultural associations. An English gloss for a Setswana zero-derived form may therefore be a workable equivalent without having the same syntax as the source. The Australian reference market—specialist digital subscriptions, university and government projects, public-library access and community partnerships—can support a bilingual entry, but it still needs language-specific description rather than an imported English template.

Editing, publishing, and maintaining the record

Editing should move from evidence to decision. A useful worksheet records each form, proposed word classes, citation, distribution, syntax, semantic relationship, historical note and reviewer verdict. Two Setswana-speaking reviewers should examine difficult cases independently before an editor resolves them. The style guide should define terms such as conversion, zero derivation, deverbal noun, homonym and lexicalised use, then provide examples showing how the labels differ. This internal consistency matters when freelance contributors, students and community experts prepare entries together.

The published record should show uncertainty without making the entry cumbersome. A short sense note can explain that a noun use is established in speech but rare in edited prose, or that a form appears mainly in one region. Online editions can carry fuller citations, variant spellings and links to archived examples, while print remains synchronised with the same version. Maintenance plans should identify who checks new corpus evidence, how corrections are dated and how community feedback is acknowledged. A dictionary that presents itself as a fixed authority will age badly; one designed as a revisable scholarly record can remain useful.

The immediate next step is a controlled pilot: select twenty candidate noun–verb pairs from an existing Setswana corpus, classify them with the worksheet, obtain independent review from two Setswana-speaking linguists and record every disagreement before choosing entries for publication.