The ole' Barebones is close to that 
    http://www.tei-c.org/Vault/Bare/#ID11.1

11.1 Elements in the Bare Bones Tag Set

The tags included in the Bare Bones TEI Subset are:

    * TEI header tags (not explained; use by rote)
          o teiHeader
          o fileDesc
          o titleStmt
          o title
          o publicationStmt
          o sourceDesc
    * Paragraphs and Other Chunk-Sized Elements
          o p
          o note
          o list, item, head
    * Verse and Drama
          o l
          o lg
          o sp
          o stage
    * Miscellaneous
          o hi
          o q
          o ref
          o pb
          o gap
    * Bibliographic References
          o bibl
          o title
    * Overall Text Structure
          o tei.2
          o text
          o front
          o body
          o back
          o div, head
    * Title Pages
          o titlePage
          o docTitle, titlePart
          o docAuthor
          o docDate
          o docImprint

But be very careful about what you ask for, lest you join me 
in being admonished with "every malediction in the book" for such 
"craziness"[1]  :-)  Maybe the TEI should create an anonymized 
suggestion box?  

M

[1] http://www.tei-c.org/Vault/Bare/#ID1



>> From [log in to unmask] Mon Feb 27 09:28:53 2006
>> Date:         Mon, 27 Feb 2006 10:27:27 -0500
>> From: "Paul F. Schaffner" <[log in to unmask]>
>> Subject: Re: TEI Lite for P5
>> Comments: To: Martin Mueller <[log in to unmask]>
>> To: [log in to unmask]
>> 
>> 
>> I'm with Edward Vanhoutte on this point, but would go a little further. For 
>> pedagogical purposes, TEI-Lite isn't light enough: you'll want a barebones 
>> tag set with somewhere between the three and five dozen tags most widely used 
>> for materials that are not particularly abstruse.
>> 
>> Those numbers correspond to the size of our *production* tagsets. 
>> Production tagsets are of course always at least partly pedagogical
>> (since, given turnover,  some portion of the staff is always in
>> the novice class); but there are more important reasons for us
>> to stick with such constrained tagsets: especially efficiency of editing 
>> and review, and feasability of supervising an outsourced keying-and-coding
>> operation. Many of the features specified by the elements that we
>> omit are features that we do not believe outsourced coders can be
>> trusted to recognize consistently and/or features that we do not
>> consider worth capturing (in terms of cost and benefit).
>> 
>> Purpose/project             Element count
>> -----------------------------------------
>> Middle English Dictionary   30
>> Knight's Mechanical Dict.   30
>> TCP (EEBO/ECCO/Evans) txts  60
>> Generic dtd for vendor use  80
>> TEI-lite and derivatives    140
>> 
>> With the TCP texts, even the 60 elements may be misleading. Omitting
>> rarely used tags (LB ADD UNCLEAR) and metadata tags, and collapsing the
>> numbered DIVs to DIVx leaves about  forty tags that account for
>> 99%+ of the actual text encoding:
>> 
>> DIVx     HEAD     TRAILER     OPENER     CLOSER     SIGNED     SALUTE
>> ARGUMENT     BYLINE     POSTSCRIPT     BIBL     Q     NOTE     EPIGRAPH
>> DATELINE     DATE     P     L     LG     FIGURE     MILESTONE     PB
>> LIST     ITEM     LABEL     TABLE     ROW     CELL     FRONT     BODY
>> BACK     TEXT     GROUP     GAP     SP     SPEAKER     STAGE     REF
>> PTR     HI
>> 
>> I won't claim I don't sometimes miss the others, but less often that you
>> might suppose.
>> 
>> pfs
>>