The Enemy from Within: A Study of Political Delegitimization Discourse in Israeli Political Speech

Naama Rivlin-Angert, Guy Mor-Lan · EMNLP 2025 · 2025 · political-science · record rivlinangert2025-enemy-within
tag layer · index · concepts · authors · עברית

Where this record's text came from. faceted 2026-08 by reading the locally held text (pdftext); every evidence string verified as an exact substring of that text before merging. Claim type is method: although the paper reports substantive empirical findings (a rise in PDD across three decades, platform, gender and bloc differences), it frames its contribution as a new task definition, corpus and classification pipeline. method is system-or-tool-building for the same reason — the deliverables are an annotation scheme, an annotated corpus and benchmarked encoder/decoder classifiers — but the analysis section is essentially observational corpus analysis with significance tests, which the method vocabulary has no term for. Replication status is untested; the paper states no computational framework for PDD previously existed ("To the best of our knowledge, no computational framework has yet been developed to identify and track PDD systematically over time or across political actors."), so there is nothing to replicate and no replication attempt is reported. One further relation (this paper filling the gap left by prior manual-coding studies of Berrocal 2019, Baldi and Franco 2015 and others) was deliberately omitted: the only sentence stating it contains U+037E characters in place of the semicolons separating the citations, a PDF-extraction artifact I did not want to risk mis-transcribing. The Limitations section of this source text is truncated mid-sentence after "For example," so any further stated limitations were unavailable. 2026-08-14, 2026-08-16 and 2026-08-25 schema-0.3 passes: every quote verified as an exact substring of the staged text and anchored at merge time; concept terms verified present; every edge endpoint an existing node.

How to read this page. The article itself is not reproduced here — this paper is not cleared for redistribution, so what follows is the tag layer alone. The evidence sentences are withheld with the article. The table at the foot lists every tag the record carries; each row names the facet it belongs to, and the whose words column says whether the tag records the paper's own wording or the tagger's sentence about the paper. Every tag links to each other paper in the corpus carrying it. The machine-readable edition of the same thing is the TEI file beside this one, where the tags live in standOff, and the article is at its publisher via the link below. When we have changed our mind. Some entries in the table begin with a word in capitals and the reason for it. EVIDENCE CORRECTED — the finding still stands, but we had quoted the wrong sentence from the paper, and the quotation has been replaced with the right one. RETRACTED — the entry itself was wrong; it stays on the page, marked, with the reason, rather than being deleted. WITHDRAWN AND REPLACED — the entry was wrong, and a corrected version of it appears in the entries below. Each mark carries its own reason, in brackets, because the reasons differ from one entry to the next. An entry we have WITHDRAWN is removed from this page and kept in the record, with its reason. A line reading Audit and a date is a later re-reading of the entry against its whole sentence: DOUBTFUL means it was kept, and the reason says what is loose about it. Quotation marks inside the tagger's sentence. Where a tag is the tagger's own sentence, a phrase in quotation marks inside it is borrowed from the paper — the tagger's wording around it is not. The paper's own sentence is the evidence quote for that row; on this page it is withheld with the article, and the article is at its publisher. Nothing inside quotation marks anywhere on this page is ours: if it is quoted, it is the paper's. Whose words a concept carries. A concept entry records a term the paper itself uses, so the term, the sense given for it, and its evidence are the paper's own words, quoted where the paper fixes them. This is the opposite of a sense entry in the tag table, which is the tagger's sentence about how the paper uses a term. Where a concept entry could not be grounded in the paper's words, the entry says so on its own row.

Other fields define these differently the same word, defined another way in another discipline
classification 2 fields · polarization 2 fields

Tags on this paper

Every tag the record carries, each linking to every other paper in the corpus that carries it. A tag appears here whether or not its words occur in the article — the spans below are only the ones a sentence in this paper licenses. Of the 10 concept tags on this paper, 5 are grounded in a quoted sentence.

annotation political delegitimization text classification affective polarization large language model Political Delegitimization Discourse PDD policy critique incivility annotation scheme a two-stage pipeline: a binary classifier first decides whether a sentence carries PDD at all, and only sentences it marks positive go on to be classified for properties and target spans the category is drawn by what the utterance DOES to the target's standing rather than by whether it is rude, and the paper says this is why classifiers built on surface hostility miss the subtle cases the boundary against policy criticism is enforced by a worked NEGATIVE example rather than by a rule every proportion is computed by aggregating at the SPEAKER level first, so that prolific politicians cannot dominate the rate distinctive targets per political bloc are identified with a weighted log-odds method rather than by raw frequency delegitimization classification parliamentary speech polarization Hebrew corpus inter-coder reliability on a 170-sentence sample: average correlation 0.91 and average Cohen's Kappa 0.82, called substantial agreement the rise in parliamentary PDD: mean share 6.6% to 9.8%, reported with t = -105.24 and p < 0.01 alongside |d| = 0.13 the gender difference: 0.066 against 0.055, called modest in the same sentence as significant, t = 2.613, p = 0.010, |d| = 0.268 the corpus: 10,410 annotated sentences, 1,812 PDD-positive at 17.4%, and 642 carrying the second-stage annotations system-or-tool-building

Legend

relation to other work	`claim_relations`	how this paper stands to another: agrees, contradicts, extends, qualifies
claim type	`claim_type`	the shape of the claim: existence, mechanism, method, critique, prediction, null
concept edge	`concept_edges`	an asserted relation between two concepts, with the sentence asserting it
concept	`concepts`	an idea the paper works with, quoted where the paper uses it
content tag	`content_tags`	a finding, definition, quantity, statistic or variable named in the paper
control	`controls`	an experiment run to eliminate a rival explanation
limitation	`limitations`	what the paper says its evidence cannot support
mechanism	`mechanisms`	how the reported effect is said to come about
method	`method`	how the paper produced its evidence
phenomenon	`phenomenon`	the effect or regularity under study
sense	`senses`	a distinct use of a term, quoted where the paper fixes it
transfer	`transfers`	a method or concept carried in from another field

The paper

The full text is not reproduced here. This article is not cleared for redistribution, so the page carries the tag layer only. Its evidence sentences are withheld; the grounding status remains with each tag, and the article itself is at its publisher. The reading view exists for the papers whose licence permits it.

Every tag on this page

claim_relations 4 · claim_type 1 · concept_edges 2 · concepts 5 · content_tags 6 · controls 1 · limitations 8 · mechanisms 5 · method 1 · phenomenon 1 · senses 3 · transfers 2

Relations on this page carry a quotation for the claim. Where a relation was separately evidenced — and so tested against the vocabulary for its type — the row says so; none of 2 relations here carry that mark. The rest are the tagger’s reading of the sentence quoted beside them.

facet	tag	field	conf.	whose words
claim_type	method Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_type`		the paper's words
phenomenon	political delegitimization discourse Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`phenomenon`		the paper's words
method	system-or-tool-building Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: system-or-tool-building	`method`		the paper's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: a two-stage pipeline: a binary classifier first decides whether a sentence carries PDD at all, and only sentences it marks positive go on to be classified for properties and target spans	`mechanisms[0]`	0.9	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the category is drawn by what the utterance DOES to the target's standing rather than by whether it is rude, and the paper says this is why classifiers built on surface hostility miss the subtle cases	`mechanisms[1]`	0.9	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the boundary against policy criticism is enforced by a worked NEGATIVE example rather than by a rule	`mechanisms[2]`	0.9	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: every proportion is computed by aggregating at the SPEAKER level first, so that prolific politicians cannot dominate the rate	`mechanisms[3]`	0.9	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: distinctive targets per political bloc are identified with a weighted log-odds method rather than by raw frequency	`mechanisms[4]`	0.9	the tagger's words
limitations	The study's scope and transferability are limited. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`limitations[0]`		the tagger's words
limitations	Findings from the Hebrew-language, Israeli case may not generalize to other languages or political systems. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`limitations[1]`		the tagger's words
limitations	The social media analysis covers only Facebook posts from 2018-2021, excluding other platforms and later periods. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`limitations[2]`		the tagger's words
limitations	Porting the approach to other languages and political contexts is expected to be difficult because delegitimization is shaped by local dynamics. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`limitations[3]`		the tagger's words
limitations	The category is defined by excluding policy criticism, a line a coder must draw in each sentence. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: any count of delegitimizing speech from this corpus: a sentence counts only if the hostile characterization is Quotation withheld, which is a judgement made per sentence	`limitations[4]`	1.0	the tagger's words
limitations	Reliability was established on a 170-sentence sample by coders who include the study's lead author. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: the reliability of the labels underlying these results: annotation was by three human annotators including the lead author, with agreement measured on a sample of 170 sentences	`limitations[5]`	1.0	the tagger's words
limitations	Only a sixth of the labelled corpus carries the finer annotations the characterisation rests on. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: claims here about intensity, incivility, target type or affective framing: those annotations exist for 642 instances, against 1,812 sentences exhibiting the phenomenon in a corpus of 10,410	`limitations[6]`	1.0	the tagger's words
limitations	The category is proposed by this paper rather than inherited from the literature. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: treating PDD as an established construct: it is proposed in this paper as a distinct analytical category alongside the pipeline built to detect it	`limitations[7]`	1.0	the tagger's words
transfers	transformer encoder models (multilingual mBERT, and the Hebrew-targeted AlephBERT and HeRO) and LLM decoders, applied to detect political delegitimization discourse in Knesset and social media text Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`transfers[0]`	0.8	the tagger's words
transfers	the weighted log-odds method for identifying tokens most strongly associated with a group, applied to political blocs Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`transfers[1]`	0.75	the tagger's words
controls	the annotated dataset split 70 per cent train, 15 per cent validation, 15 per cent test Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. eliminates: reporting classifier performance on data used to fit or to select it - the three-way split separates fitting, model choice and reporting · outcome (stated by the tagger, no quotation): held	`controls[0]`	0.85	the tagger's words
concepts	Political Delegitimization Discourse Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: Political Delegitimization Discourse	`concepts[5]`		the paper's words
concepts	PDD Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: PDD	`concepts[6]`		the paper's words
concepts	policy critique Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: policy critique	`concepts[7]`		the paper's words
concepts	incivility Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: incivility	`concepts[8]`		the paper's words
concepts	annotation scheme Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: annotation scheme	`concepts[9]`		the paper's words
senses	PDD - A STIPULATED CATEGORY WHOSE BOUNDARY IS SYMBOL VERSUS POLICY: hostility toward a political actor counts only when it is not grounded in policy critique, so the same anger about the same target falls inside or outside the category depending on what the coder judges it to be aimed at Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[0]`	1.0	the tagger's words
senses	the boundary, SHOWN RATHER THAN STATED: the paper prints a sentence calling a budget an example of negligence and promiscuous conduct endangering the economy and society, and records that it is NOT annotated as PDD - a published negative case, which is rarer in this corpus than a definition Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[1]`	1.0	the tagger's words
senses	the conceptual definition BEHIND the annotation scheme: discourse aimed at discrediting political groups or actors by attacking their symbolic and affective dimensions - the scheme is built on that phrase, and every count in the paper inherits it Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[2]`	1.0	the tagger's words
claim_relations	builds-on → Rivlin-Angert (2025), on delegitimization of political identities Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[0]`		the tagger's words
claim_relations	builds-on → Mor-Lan et al. (2024), IsraParlTweet dataset of Knesset speeches Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[1]`		the tagger's words
claim_relations	builds-on → Monroe et al. (2008), weighted log odds method Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[2]`		the tagger's words
claim_relations	extends → existing NLP tasks: toxicity detection, stance detection, targeted sentiment analysis Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[3]`		the tagger's words
concept_edges	PDD —same-as→ Political Delegitimization Discourse Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. strength of the relation: 1 also in: PDD · Political Delegitimization Discourse	`concept_edges[0]`		the tagger's words
concept_edges	PDD —distinct-from→ policy critique Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. strength of the relation: 1 also in: PDD · policy critique	`concept_edges[1]`		the tagger's words
content_tags	Delegitimization intensity was rated on a three-point scale. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: QUOTE RECOVERED FROM THE PAPER'S OWN CHARACTERS. The recorded quote was LETTERS-ONLY IDENTICAL to this stretch of the staged text but not an exact substring of it - spaces, hyphens, ligature glyphs or mathematical operators differed, so it could never anchor. This is the file's own text between the same two letters: not produced by a rule about the damage, copied out of the file. Round-trip checked (the recovered span is letters-only equal to the original) and length-checked. The text as originally recorded is kept in evidence as extracted.	`content_tags.findings[0]`		the tagger's words
content_tags	Early computational approaches used supervised classifiers to detect abusive language aimed at politicians. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`content_tags.findings[1]`		the tagger's words
content_tags	inter-coder reliability on a 170-sentence sample: average correlation 0.91 and average Cohen's Kappa 0.82, called substantial agreement Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: inter-coder reliability on a 170-sentence sample: average correlation 0.91 and average Cohen's Kappa 0.82, called substantial agreement	`content_tags.statistics[0]`	0.9	the tagger's words
content_tags	the rise in parliamentary PDD: mean share 6.6% to 9.8%, reported with t = -105.24 and p < 0.01 alongside \|d\| = 0.13 Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the rise in parliamentary PDD: mean share 6.6% to 9.8%, reported with t = -105.24 and p < 0.01 alongside \|d\| = 0.13	`content_tags.statistics[1]`	0.9	the tagger's words
content_tags	the gender difference: 0.066 against 0.055, called modest in the same sentence as significant, t = 2.613, p = 0.010, \|d\| = 0.268 Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the gender difference: 0.066 against 0.055, called modest in the same sentence as significant, t = 2.613, p = 0.010, \|d\| = 0.268	`content_tags.statistics[2]`	0.9	the tagger's words
content_tags	the corpus: 10,410 annotated sentences, 1,812 PDD-positive at 17.4%, and 642 carrying the second-stage annotations Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the corpus: 10,410 annotated sentences, 1,812 PDD-positive at 17.4%, and 642 carrying the second-stage annotations	`content_tags.statistics[3]`	0.9	the tagger's words

Interpretive choices

Tags that could not be placed in the text:

none — every tag was located in the text

What was deliberately not tagged, and why, is recorded per paper by the tagger. Where a paper has no such record, this page says so rather than implying that nothing was declined.