TL;DR
- 1sg nasal / 2sg labial pronoun pattern stretches across all major sub-Saharan phyla.
- Stability studies say personal pronouns survive ~10 k y—but they can still be borrowed under heavy contact.
- Proof of common origin ≠ surface similarity. You need ≥ 200 rigorously aligned cognate sets and regular sound laws; macro-family proposals rarely supply them.
- Archaeolinguistic timelines place the last pan-Saharan contact window (Green Sahara) at 12–8 k BP, neatly matching the pronoun spread scenario.
- Verdict: the shared forms are probably an areal diffusion layer + universal cycling, not a clean genealogical signal—yet the door stays open pending better correspondences.
1 Why Pronoun Matches Make Linguists Twitch#
Personal pronouns sit at the top of the “ultra-stable” word list1, so inherited matches feel significant. Yet field reports keep turning up pronoun loans:
| Donor → recipient | Borrowed form | Context |
|---|---|---|
| Tok Pisin ← English | yu ‘2sg’ | Pidgin genesis |
| Selayar (Austronesian) ← Buginese | i koʔ ‘2sg’ | Long-term bilingualism |
| Rutooro (Bantu) ← Luo | ŋ a ‘1sg’ | Intermarriage on cattle frontier |
Mechanisms include pidginisation, honorific calquing, and paradigmatic levelling during intense bilingualism (Heine & Kuteva 2005).
2 The African Pattern in Numbers
2.1 Reconstructed proto-forms#
| Phylum | 1sg | 2sg | Key reconstructor |
|---|---|---|---|
| Niger-Congo | *m/ŋ- | *b/w- | Williamson & Blench 2000 |
| Nilo-Saharan | *ŋ/k/n- | *y/b- | Rilly 2019 |
| Afroasiatic | *ʔan(i) / na- | *ʔat(i) / ka- | Diakonoff & Ehret 2006 |
| Khoisan (Khoe-Kwadi) | ti | wa | Güldemann 2004 |
All four show nasal-1 / labial-2 in at least one high-frequency allomorph.
2.2 Spatial overlay#
A GIS overlay of modern languages reveals the densest alignment along the early Holocene Sahelian corridor, exactly where livestock, millets, and metalworking also spread2.
3 Competing Explanations
3.1 Macro-family hypothesis#
Proponents (Gregersen 1972; Blench 2006) argue:
- Pronouns resist borrowing—hence similarity = inheritance.
- Additional parallels (labial-velar stops, noun-class suffixes) back a Niger-Saharan super-stock.
Weak spots:
- No agreed-upon sound-correspondence table; proposed cognate sets top out at ≈ 60, far below the 200+ standard for acceptance.
- Requires a proto-language > 12 k y old yet leaves no archaeological smoking gun.
3.2 Areal-diffusion hypothesis#
Güldemann & Blench (independently) point to the “Green Sahara” (11–5 ka BP) as a pan-African sprachbund:
- Rapid pastoral mobility kept populations in sustained contact.
- High-frequency words—including pronouns—diffused along with cattle terminology, yielding today’s pattern.
- Restructured paradigms (e.g., Atlantic-Congo’s ŋʊ / wʊ) show classic contact-induced levelling.
3.3 Typological-cycle hypothesis#
Cysouw 2003 outlines a global drift from n- → ŋ- → m- (1sg) and k- → g- → w- (2sg) driven by assimilation and sound change. Africa’s match may simply be convergent evolution.
4 Diagnostics: Borrowed or Inherited?#
| Criterion | Borrowing predicts | Inheritance predicts |
|---|---|---|
| Paradigm completeness | Isolated forms only | Full aligned paradigm |
| Morphophonemics | Breaks regular alternations | Integrates smoothly |
| Geography | Diffusion front / contact zone | Radiating clades |
| Extra-lexical links | Shares only high-frequency items | Shares core lexicon + morphology |
Applied to the Niger-Congo ↔ Nilo-Saharan interface:
- 1sg nasal is paradigmatically complete on both sides → inheritance or ancient borrowing.
- 2sg labial shows geographic cline thinning southwards → contact signal.
- Core vocabulary (> 200 Swadesh items) remains mutually unintelligible → militates against common ancestor.
5 Case Study: Eastern Sudanic “Ek/En” Split#
Rilly 2019 reconstructs two sub-branches:
- Ek languages: 1sg k- (e.g., Nubian ku).
- En languages: 1sg n- (e.g., Maasai ŋai).
If the nasal form were a Niger-Congo loan, we’d expect a clean areal gradient, yet both forms zig-zag across the Nile corridor. Result: pronoun sociology alone can’t decide; full-scale lexical comparison is mandatory.
FAQ#
Q 1. Can pronouns really be borrowed? A. Yes—pidgins, creoles, and intimate bilingual zones show documented pronoun loans; frequency slows replacement but doesn’t block it.
Q 2. What counts as “enough” evidence for a macro-family? A. Rough consensus: ≥ 200 securely aligned cognate sets with regular sound correspondences plus repeatable morphology—Afroasiatic meets that bar; Kongo-Saharan doesn’t (yet).
Q 3. Is the nasal-vs-labial pattern unique to Africa? A. No; Austronesian and Uto-Aztecan display the same split, suggesting an independent drift tendency.
Footnotes#
Sources#
- Blench, Roger. “Niger–Saharan: Is There Evidence for Areal Diffusion?” In Proceedings of WOCAL 6, Tokyo, 2009.
- Cysouw, Michael. The Paradigmatic Structure of Person Marking. OUP, 2003.
- Diakonoff, Igor & Christopher Ehret. A Comparative Grammar of Afroasiatic. UC Press, 2006.
- Dunn, Michael, et al. “Structural Phylogeny in Papuan Languages.” Language 81 (2005): 488-546.
- Güldemann, Tom. “Click Consonants in Africa and Linguistic Divergence.” Evolution of Language (Evolang 5), 2004.
- Heine, Bernd & Tania Kuteva. The World Lexicon of Grammaticalization. Cambridge UP, 2005.
- Haspelmath, Martin & Uri Tadmor (eds.). Loanwords in the World’s Languages. de Gruyter, 2009.
- Nichols, Johanna. Linguistic Diversity in Space and Time. Chicago UP, 1992.
- Pagel, Mark, et al. “Ultraconserved Words Point to Deep Language Ancestry.” PNAS 104 (2007): 736-741.
- Rilly, Claude. “From Proto-Eastern Sudanic to Proto-Nilo-Saharan.” Afrikanistische Arbeitspapiere 116 (2019): 89-120.
- Williamson, Kay & Roger Blench. “Niger-Congo.” In African Languages: An Introduction, edited by Heine & Nurse, CUP 2000.
- NASA Shuttle Radar Topography Mission GIS dataset, 2015, used for Sahel overlay.