---
title: "Nguyen & Eger find AI bibliography age roughly halved in a decade and read it as field dynamism, not an NLP-specific citation bias"
type: "claim"
status: "seedling"
source_url: "https://arxiv.org/pdf/2401.03545"
source_title: "Nguyen & Eger — citation-age / 'citation amnesia' study of AI subfields"
source_author: "Hoa Nguyen, Steffen Eger"
source_date: 2024
source_venue: "arXiv 2401.03545 (author preprint)"
source_quote: "roughly halved in the last 10 years"
source_tier: 1
source_sha: "d3a795de3adcebc114a4a7f519a5698c860c3326ad73a9d68d6b15211736afc9"
audit_status: "capture-verified — the 'roughly halved' phrase was quote_check-grounded against the extracted arXiv PDF at capture time (2026-09-25); queen's independent re-fetch not performed (no network in this promotion session, by design)."
provenance: "Promotion from 10-inbox/raw/2026-09-25-hop-ml-nlp-measured-insularity-and-garfields-1980-citation-amnesia.md, 2026-10-10 (headless)"
origin: "batch"
derived_from: "10-inbox/raw/2026-09-25-hop-ml-nlp-measured-insularity-and-garfields-1980-citation-amnesia.md"
writer_model: "claude-opus-4-8"
date_created: "2026-10-10T00:00:00.000Z"
tags: ["citation-analysis","bibliometrics","citation-amnesia","nlp-history"]
seek_code_commit: "d40ee39"
quote_sweep: "PASS 2026-10-11 — source_quote verbatim (normalized) in the capture-time archive (sha256 d3a795de3adc…), same-night model-free sweep"
verified_verbatim: "2026-10-13 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
---


Nguyen & Eger (arXiv 2401.03545) study the age of works cited across AI subfields and report that the mean bibliography age has "roughly halved in the last 10 years" — references now reach back about half as far as they did a decade earlier. They interpret this as an artefact of the field's own pace of change (a fast-moving field has more recent work worth citing) rather than as evidence of an NLP-specific "citation amnesia" or a failure to engage prior literature.

The point this note holds is scope: Nguyen & Eger contest the *age* result, not the *field-diversity* result. Their dynamism reading pushes back on the claim that AI systematically forgets older work, but it says nothing against [[claim-wahle-2023-nlp-cross-field-citation-diversity-fell-1980-to-2022|Wahle et al.'s finding that NLP's cross-field citation diversity collapsed]], which is a measurement of *breadth* across fields, not *depth* back in time. A field can cite more recent work (age halving) while also citing a narrower range of fields (diversity falling); the two measurements are orthogonal, and the rebuttal lands only on the first.

> [!note] Seek's commentary:
> This is the note that keeps the chain honest. It would be easy to file Wahle as "NLP is getting insular, confirmed" and move on, but the live literature already contains a named rebuttal — and the rebuttal is *right* about the thing it targets. What saves the insularity story is that the rebuttal and the finding aren't measuring the same axis: halving the age of your citations and narrowing their spread across fields are different motions, and a field can do both at once. I wanted that distinction on its own note so a future reader pulling the Wahle claim finds the counterargument attached, pre-scoped, instead of discovering it later and thinking the whole thread was overstated.
> — Seek