What Word Embeddings Show About Xenophon

Three previous notes in this series have traced the semantic neighborhood of βάρβαρος across Herodotus, Euripides, and Plato, watching the term shift from a military label (Herodotus) to a political-social category (Euripides) to a consolidated dyad with Ἕλλην (Plato). The Xenophon model, trained on the same surface-form fastText framework, completes the four-author comparison and produces the most distinctive profile of the series: βάρβαρος in Xenophon is not simply a regression to Herodotus’s military cluster. It has its own shape.

The vocabulary is 10,126 tokens — the smallest of the four models. βάρβαρος’s nearest neighbors are πολέμω (to make war, 0.552), κακόνοος (malevolent, ill-disposed, 0.491), ἀναμίγνυμι (to mix or mingle, 0.442), and Τισσαφέρνης (0.417) — the Persian satrap who betrayed the Greek commanders at the opening of the Anabasis. A specific named enemy is the fourth nearest neighbor of the generic ethnic label. This is a precise summary of what the Anabasis does: it puts Barbarians in individual, dramatic relation to the Greek expeditionary force, named and identified, rather than treating them as an undifferentiated mass. And the nearest neighbor of Ἕλλην itself is Ἀριαῖος (0.511) — the Persian commander who fought alongside the Ten Thousand at Cunaxa and then negotiated his own treaty with the King, abandoning the Greeks. Greek identity in Xenophon’s military corpus is triangulated through its relationship with specific barbarian individuals, not through abstract political or philosophical concepts.

The four-author comparison table now reads:

The most striking number is τύραννος at −0.009 — essentially zero, and slightly negative. Euripides and Plato both score above 0.290 on this association: tyranny is one of the political concepts most strongly attached to the barbarian label in both tragedy and philosophical prose. In Xenophon it disappears entirely. The reason is the Cyropaedia. Xenophon’s life of Cyrus the Great is not a portrait of Oriental despotism; it is a sustained argument that Cyrus represents the ideal of just and effective kingship, in explicit contrast with the tyrant who rules through fear and self-interest. The Hiero, Xenophon’s dialogue on tyranny, develops the same contrast from the other direction: the tyrant is miserable, hemmed in by suspicion, deprived of the ordinary pleasures available to the private citizen (ἰδιώτης is τύραννος’s nearest neighbor in this model, at 0.528). Xenophon wrote more and more carefully about tyranny than almost any other classical author — which is precisely why the concept does not cluster with βάρβαρος. He had separated the two ideas that tragedy routinely fused.

The second distinctive feature is γένος at 0.211, the highest-scoring concept for βάρβαρος in Xenophon and the only one that rises clearly above background. Probing the full peoples matrix confirms that this is not specific to βάρβαρος: γένος scores as the highest-similarity concept for every people in Xenophon — Πέρσης (0.203), Αἰγύπτιος (0.234), Σκύθης (0.357). Ethnic identity in Xenophon is organized around descent and lineage before law, freedom, tyranny, or any other political concept. The γένος profile itself makes clear what kind of lineage is at stake: its nearest neighbors include εὐγενής (0.444) and αἷμα (0.285), and among the named terms Περσείδης (son of Perseus) appears — ancestry tracked through named progenitors, noble birth verified through bloodline. This is the aristocratic sense of γένος, not the philosophical natural-kind sense that dominates in Plato. In Xenophon, lineage is inheritance of status, not membership in a biological species or logical taxon.

φύσις in Xenophon is −0.061 for βάρβαρος and −0.096 for Ἕλλην. Both sides of the antithesis are negatively associated with nature. This is not the result of a small or noisy vocabulary — it holds consistently across all peoples in the similarity matrix. Where φύσις does appear in the text, it clusters with ψυχή (0.523), φιλότιμος (0.494 — love of honor), παιδεία (0.493), εὐφυής (0.449 — naturally gifted). This is the educational-practical register of the Memorabilia and Cyropaedia: nature as raw aptitude that good upbringing can develop into excellence. φύσις in Xenophon is a concept about individual potential and pedagogy, not about ethnic group constitution.

The result for ψυχή (−0.148) and ἀρετή (−0.106) against Ἕλλην is worth dwelling on. In none of the other authors is Greek identity negatively associated with philosophical virtue discourse. In Herodotus, Ἕλλην → ἀρετή is moderate and positive; in Plato the philosophical vocabulary is at least in a separate neutral register; in Xenophon it is mildly negative for both the Greek and the barbarian label. This is the signature of a corpus organized around military operations and practical leadership rather than philosophical inquiry into character. The Socratic dialogues Xenophon wrote — the Memorabilia, the Symposium — do not pull enough frequency weight to move ψυχή and ἀρετή into the ethnic discourse field. What dominates is the operational vocabulary of the Anabasis and Hellenica.

The γένος cross-author comparison shows where Xenophon stands relative to the others:

Herodotus: γένος as political dynasty (high τύραννος, low biological terms). Euripides: γένος as biological reproduction (high φύω and σπέρμα). Plato: γένος as ontological natural kind (high εἶδος and φύσις). Xenophon: γένος as noble birth (highest εὐγενής of any author, significant αἷμα and φύω but no dominant philosophical term). The four-author γένος trajectory does not describe a straight line from political to biological to philosophical — it describes four distinct modes of conceptualizing group membership, each calibrated to the author’s intellectual register.

The full four-author profile now permits a summary statement about per-author variation in this discourse. The political associations of βάρβαρος — tyranny, slavery, law, freedom — are not universal features of Greek ethnic discourse; they are concentrated in tragedy and Platonic political philosophy. The military-operational register is characteristic of historiography in both its archaic (Herodotus) and classical (Xenophon) forms. The one finding that holds across all four authors is the one most relevant to the naturalization hypothesis: φύσις never rises. The concept of inherent natural constitution does not attach to the ethnic label in any of these corpora, whatever the individual author’s philosophical orientation.

A fifth note will apply the same framework to CADE’s century-sliced diachronic models, which track these patterns across six centuries from the archaic period through the Hellenistic, and examine whether the per-author profiles are representative of their respective periods or idiosyncratic to individual writers.