Historians debate this endlessly, but a strong case favors artifacts—physical objects—as the more reliable primary evidence. Here's why, with caveats.
Objects are stubbornly empirical. A stone tool reveals its manufacturing technique (flaking pattern), material source (petrography), use-wear (microscopic analysis), and date (stratigraphy, radiocarbon). No scribe's bias, no translation error, no rhetorical agenda. The thing is its own witness.
Archaeology's scientific methods—thermoluminescence, dendrochronology, DNA residue analysis—extract data the maker never recorded. A potsherd tells us about trade networks, diet, technology, social organization.
Texts are mediated: authored, edited, copied, translated, censored. They reflect the writer's perspective, audience, and intent. Ancient histories (Herodotus, the Bible, Sima Qian) blend fact, myth, propaganda, and theology. Corroboration is often scarce.
Well-preserved texts provide narrative specificity artifacts cannot: names, dates, motives, speeches, legal codes, prices, treaties. The Rosetta Stone's trilingual inscription unlocked Egyptian hieroglyphs—an artifact enabling textual recovery. The Dead Sea Scrolls gave us Second Temple Judaism's own words.
Reliability depends on:
Best practice: triangulate. Artifacts ground; texts narrate. Together, they build history's scaffold.