Rethinking “Unrelated Roots”: Precision, Over-Precision, and the Case of andr- and anthrōp-

Historical linguistics has long relied on the principle of regular sound change, a methodological success that enabled the reconstruction of large parts of Indo-European and other families. Yet its strength contains a limitation: precision can drift into over-precision. By eliminating all irregular paths as “ad hoc,” one risks overstating the negative evidence—confidently declaring two forms “unrelated” even when the available evidence cannot conclusively support such a claim.

The long-standing classification of Greek ἀνήρ / ἀνδρός (andr-) and ἄνθρωπος (anthrōp-) as “completely unrelated roots” exemplifies this tension. Under the standard comparative method, the conclusion is straightforward: the expected sound developments of Indo-European h₂nḗr yield andr-, not anthr-, and there is no regular derivational pathway from one to the other. From this perspective, the two forms do not share a reconstructible Indo-European ancestor within the canonical sound laws. This statement is secure within that methodological frame.

However, the methodological frame is not the empirical universe.

Once one introduces well-attested real-world mechanisms—dialect differentiation, substrate influence, contact-induced restructuring, cluster repair strategies, or the well-documented distortions introduced by Bronze Age syllabaries—absolute negative statements become much less secure.

Several points illustrate the issue:

1. Instability of nasal + dental clusters in Greek and the Aegean. Clusters such as NT, NTH, NDR, NTR, and NTHR show recurrent divergent developments (assimilation, aspiration, metathesis, epenthesis). These transformations are not hypothetical; they are structurally regular within Greek dialect history and within the larger Aegean substrate.

2. Plausible shared preforms under dialect divergence. If one posits a pre-Greek cluster such as *antr- or *anthar-, two divergent daughter forms—one with assimilation → andr-, one with aspiration + epenthesis → anthr-—are structurally and typologically unproblematic. The comparative method cannot confirm such a proto-form, but neither can it exclude it on phonotactic grounds.

3. Contact effects and re-borrowing cycles. The Aegean–Anatolian interaction sphere shows repeated cases where final stops in borrowed words functioned as class markers or determinatives, then re-entered Greek as part of the lexical stem. Under such mechanisms, the -p- in anthrop- need not be inherited; it can arise through reinterpretation during the passage through a non-alphabetic intermediary.

4. Orthography as a shaping force, not a passive record. Linear A and Linear B routinely generate pseudo-consonants through syllabic segmentation and reanalysis. A single sequence like a-to-ro-po can later be alphabetised as anthrop- without implying that p ever belonged to the spoken proto-form. Script-induced consonants of this type appear cross-linguistically; ignoring them produces false negatives.

These mechanisms are all empirically attested, not speculative inventions. None contradict the principles of the comparative method; they operate outside its core assumptions.

The consequence is methodological rather than lexical: Stating that two forms are “definitely unrelated” exceeds what the data and the method can justify. The correct scientifically cautious formulation is more nuanced: • Under standard Indo-European sound laws, ἀνήρ and ἄνθρωπος do not derive from a single proto-form. • However, multiple historically attested mechanisms—Aegean substrate patterns, dialectal divergence of unstable clusters, contact-induced reanalysis, and orthographic restructuring in Bronze Age syllabaries—provide alternative pathways by which a shared preform could have developed into the attested Greek forms.

This does not claim that the two forms are related; it claims only that the categorical assertion of unrelatedness is methodologically overstated.

The broader lesson is general: avoiding ad hoc explanations is essential, but the refusal to acknowledge known historical forces—dialect stratification, substrate layering, contact phenomena, and script-induced restructurings—can itself generate false negatives. Precision is not achieved by ruling out every path that lies outside the comparative method’s narrowest parameters. It is achieved by recognising the full set of historically grounded mechanisms and by drawing conclusions only within the domain where evidence permits.

In that light, andr- and anthrōp- need not be declared related; they simply cannot be declared in principle unrelated.

That distinction matters.