[Mova](https://md.mova.elamur.ai/free/en/) · [this page in HTML](https://mova.elamur.ai/free/en/docs/papers/santorini-1990-part-of-speech-tagging-guidelines-penn-treebank/)

[home](https://md.mova.elamur.ai/free/en/) / [docs](https://md.mova.elamur.ai/free/en/docs/) / [papers](https://md.mova.elamur.ai/free/en/docs/papers/) / Part-of-Speech Tagging Guidelines for the Penn Treebank Project (3rd Revision)

# Part-of-Speech Tagging Guidelines for the Penn Treebank Project (3rd Revision)

**Authors:** [Beatrice Santorini](https://md.mova.elamur.ai/free/en/docs/papers/authors/beatrice-santorini/) · **Year:** 1990 · **Venue:** Technical Report MS-CIS-90-47, Department of Computer and Information Science, University of Pennsylvania **Link:** unknown (technical report MS-CIS-90-47) **License of the paper:** unknown

[**49** citations in our library](https://md.mova.elamur.ai/free/en/docs/papers/cited-by/santorini-1990-part-of-speech-tagging-guidelines-penn-treebank/) of the 140673 papers whose bibliographies we readmanual

- [Publisher's page](https://catalog.ldc.upenn.edu/LDC99T42)
- [PDF](https://catalog.ldc.upenn.edu/docs/LDC99T42/tagguid1.pdf)

**Authors (1)**

1. Beatrice Santorini

Citations and author metrics from OpenAlex (); authors with the same name are not always told apart.

## Summary

These are the annotator guidelines behind the Penn Treebank part-of-speech tag set (NN, NNS, JJ, VBD, VBN and the rest). The report defines each tag and, more importantly, gives decision procedures for the hard cases that make tagging inconsistent: adjective versus noun, past tense versus past participle, particle versus preposition, and similar confusions. The tests are mostly substitution and modification tests, for example whether a word can be intensified by "very" or can take a plural ending. Many short examples illustrate each decision. The document became the de facto reference for English PTB-style tagging and is still the authority behind the XPOS column of English UD treebanks.

## How Mova uses it

- `en/seeds/` — the English grammar seeds (short declarative rule notes with conditions, exceptions, examples and sources) cite the guidelines as the authority for PTB tag decisions; about 43 seed files reference it, mostly under `en/seeds/morph/` (noun number, verb forms, adjective degree, pronouns, word formation) plus a few in `nominal/` and `verbal/`.
- Example: [Adjective used as a noun: the rich vs the natives](https://md.mova.elamur.ai/free/en/docs/en/seeds/morph/adj-as-noun/) uses the "(very) rich" test from §4.1 to separate partial conversion ("the rich" stays JJ/ADJ) from full conversion ("natives" becomes NNS/NOUN).
- The seeds combine the guidelines with classic descriptive grammars, so a tagging rule always carries both a linguistic rationale and the PTB convention it must produce.

## Effectiveness in Mova

Reference material for rule writing: it defines the target conventions rather than a technique. Not measured separately.

---

👨‍🔬💥

Topics: [Treebanks](https://md.mova.elamur.ai/free/en/docs/papers/topics/treebanks/) · [Part-of-speech tagging](https://md.mova.elamur.ai/free/en/docs/papers/topics/part-of-speech-tagging/)

## Cite

The paper itself, not this card. BibTeX from the library's record of the work.

```
@book{santorini1990part,
  title = {{Part-of-Speech Tagging Guidelines for the Penn Treebank Project (3rd Revision)}},
  author = {Santorini, Beatrice},
  year = {1990},
  url = {https://catalog.ldc.upenn.edu/LDC99T42}
}
```

- [← previous Mapping to Declarative Knowledge for Word Problem Solving](https://md.mova.elamur.ai/free/en/docs/papers/roy-2018-mapping-declarative-knowledge-word-problem/)
- [next → Mischievous nominal constructions in Universal Dependencies](https://md.mova.elamur.ai/free/en/docs/papers/schneider-2021-mischievous-nominal-constructions-universal-dependencies/)
- [↑ back to papers](https://md.mova.elamur.ai/free/en/docs/papers/)