[NB: I'll be blogging random things about my upcoming BibleTech:2009 paper; these posts will all be available in the "bibletech" category. If you're presenting a paper at the conference, might I suggest the same practice? That way they'll all be available by a search for 'bibletech' on Technorati or some other such service. —RB]
In his book A Stylometric Study of the New Testament (amazon.com), Anthony Kenny lists 99 features that he tracked across the corpus, using them as a guide to his analysis. His feature list is based on categorization of the Friberg morphology circa 1986. I believe the Friberg has undergone significant revision since then and is considered to be in at least its second edition; perhaps even the third edition. Kenny also includes some stock lexical items such as conjunction instances, preposition instances, and some specific words (e.g. θεος, λεγω). Note that Kenny did all of his counts by hand, from "the microfiche concordance ot the machine-readable version of the Analytical Greek New Testament"! (Kenny, "Note on Sources") He used a TI 58 statistical calculator for his numbers, also "the ICL 2988 machine in the Oxford University Computer Services". (Kenny, "Note on Sources").
Right now, I'm thankful for fast computers, XML and for Perl and/or C# (haven't figured out which language I'll use for the code yet).
In my paper for BibleTech:2009, I'm proposing to carry out a similar analysis, only of the LXX, using the Logos Morphology. There are several of the 99 categories that can be re-used (81, to be exact). Friberg has much more going on in adjectives, adverbs, and conjunctions than the Logos LXX Morphology; this accounts for much of the difference.
However, I think I'll be able to track up to 106 features, and perhaps more. How? Kenny did very little with participles, and even less with pronouns. I have no idea why he did little with participles because the Friberg morphology is rich in this area (even differentiating, at the 'mood' slot, between 'participle' and 'participle (imperative sense)', Kenny pp. 10-11, figure 2.1). It may have been because it would be too tedious (recall he counted by hand). But pronouns are simply a type of noun in this edition of Friberg, so Kenny's hands were tied (he tracked third-person pronouns in sum and also by case, but that's it).
Kenny also didn't track instances of the vocative case (for articles, nouns, and adjectives). But he did track optatives and pluperfects (indeed rare cases in the NT). Thus, to the 81 shared criteria, I'm considering adding 25 more for a total of 106.
If you're interested, the list of 25 additional features is below.
Because of differences in classification of voice
82. Number of occurrences of third-person singular indicative verbs in the either-middle-or-passive voice
83. Number of occurrences of verbs in the participle mood
84. Number of occurrences of participles in the nominative
85. Number of occurrences of participles in the dative
86. Number of occurrences of participles in the genitive
87. Number of occurrences of participles in the accusative
88. Number of occurrences of participles in the masculine
89. Number of occurrences of participles in the feminine
90. Number of occurrences of participles in the neuter
91. Number of occurrences of participles in the singular
92. Number of occurrences of participles in the plural
93. Number of occurrences of Proper Nouns
94. Number of occurrences of Interjections (I)
95. Number of occurrences of vocative articles
96. Number of occurrences of vocative nouns
97. Number of occurrences of vocative adjectives
Other Pronoun Information
98. Number of occurrences of Relative Pronouns
99. Number of occurrences of Reciprocal Pronouns
100. Number of occurrences of Demonstrative Pronouns
101. Number of occurrences of Correlative Pronouns
102. Number of occurrences of Interrogative Pronouns
103. Number of occurrences of Indefinite Pronouns
104. Number of occurrences of Reflexive Pronouns
105. Number of occurrences of Possessive Pronouns
106. Number of occurrences of Personal Pronouns
There may be more, I just have to think about it a bit more. For instance, I could add case-specific instances of each pronoun type (so, relative pronouns in the nominative, in the genitive, etc.) but at present I'm thinking that's overkill. Of course, I may change my mind. I also need to consider if there are particular word instances to include in the feature list; I may have to do some word frequency analysis in order to determine candidates. I will also have to review LXX-specific conjunctions and prepositions to determine how those portions of the list might be expanded.