Moby Part of Speech List — Edition Insights

The Documentation as a User Manual

The text begins with a “Quick Start” section that assumes a technical audience: “Insure you have at least 3Mb of free disk space,” it advises, and instructs users to “Create a directory to hold these files.” This is not a book to be read linearly but a resource to be extracted and deployed. The documentation emphasizes that the second edition is “a particularly thorough revision,” with “fifteen thousand new entries” and many more “scrutinized for correctness and modernity.” It claims this is “unquestionably the largest P-O-S list in the world,” a boast that sets expectations for comprehensiveness. The user is guided to delete the original zip file to save space, a practical tip that underscores the file's nature as a tool rather than a text.

Decoding the Legend

The legend explains the part-of-speech codes: each entry consists of a word or phrase followed by a backslash and a code using ASCII symbols, where “case is significant.” For example, “N” stands for noun, “p” for plural, “h” for noun phrase, “V” for verb (usually participle), “t” for transitive verb, “i” for intransitive verb, “A” for adjective, “v” for adverb, “C” for conjunction, “P” for preposition, “!” for interjection, “r” for pronoun, “D” for definite article, “I” for indefinite article, and “o” for nominative. The documentation notes that “-ed, -ing, -ly, and -ic forms” are often not explicitly listed; participle forms are marked simply with “V.” This coding system is the key to using the list, and the legend is essential for interpreting entries.

Sample Entries and Their Implications

The sample entry “engineer\Nt” is explained: the word “engineer” has two main uses, as a noun and as a transitive verb. The documentation provides example sentences: “That engineer could write in microcode” (noun) and “We sure engineered that software to death” (verb). This illustrates how a single entry can encode multiple parts of speech. The documentation also notes that words like “be” have “one listing with all the parts-of-speech for all senses concatenated.” Foreign words “commonly used in English” include diacritical marks, such as the acute accent e (ASCII 142). These details reveal the list's ambition to cover the full range of English usage, including inflected forms and borrowed terms.

Reading Between the Codes

The list's structure invites a reader to consider the relationship between word forms and grammatical functions. The documentation states that “the many included phrases means that parsing algorithms can now tokenize in units larger than a single word,” hinting at the list's utility for natural language processing. A first-time user might start by scanning the legend, then examine a few entries to see how codes are applied. The list is not meant to be read cover-to-cover but queried or processed programmatically. The documentation's emphasis on “correctness and modernity” suggests that the list has been curated, but the user must rely on the codes to interpret each entry. The lack of definitions means the list is purely a grammatical index, not a dictionary.

For a first reading, approach the Moby Part-of-Speech II as a reference work: begin with the documentation to understand the coding system, then sample a few entries to see how the codes capture grammatical behavior. The list's value lies in its scale and consistency, not in any narrative or argument. Readers interested in English grammar or computational linguistics will find it a useful dataset, but those seeking definitions or usage examples will need to look elsewhere.

Some lists, like the one peering into Grady Ward's careful tagging of every word, feel like the quiet cataloguing of a mind—each entry a small, patient gesture. I am reminded of an old bookseller's catalogue from May 1883, which whispers the same reverence for order, for the sheer pleasure of naming things. Chatto & Windus's List of Books, May 1883 — Context and Discussion shares that gentle, methodical soul. Both are maps of a vast, beloved wilderness.