Guide to the Unmerged

File last modified
For other MEC files, see the  MEC Index page.

Typical Problems Arising from MED->HB Mapping

IN GENERAL:

normal stencils get mapped from MED to HB using <grp>.
"v.d." stencils get mapped from MED to HB using <prp>.
"-----" stencils get mapped from MED to HB using <nrpgrp>.

SPECIAL PROBLEMS

1. Short/Full stencil problem:
   MED has full stencil, HB has short stencil only.

   Make list of such stencils in your batch. Add full stencils to the HB file. Add exactly the same stencil, inside grp tags, to the appropriate spot in the map file. Cut and paste the med stencils from the med file into the new grps in the map file.

2. Dating discrepancies, e.g.:
   • HB has v.d. but MED has v.d. and (v.d.)

     check (v.d.) exx. to see if legit. If so, change hb; if not, change MED. In many cases, assume that the MED is wrong, especially if the exceptions are very rare.

   • HB has (v.d.) but MED has v.d. and (v.d.)

     Same as above.

   • HB has v.d.(v.d.) but MED has v.d., (v.d.), and v.d.(v.d.)

     accept MED stencils as they are.

   • HB has single date but MED has v.d.

     if MED variation is easily accounted for typographically (e.g., 1389 for 1398, c1398 for ?c1398), change MED.

     if MED variation is more difficult to explain, check to see if the exx. are legit. If so, change HB.

     IMPORTANT: if MED variation can be explained by some *other* error, e.g., in version number or manuscript, check exx. to see if they are legit. If so, change HB; if not, fix MED. For example, a citation of ?c1425 Chauliac(1) may be an error in the date (should be ?a1425), or it may be an error for ?c1425 Chauliac(2); Lydg.TB quots. with the wrong date may belong to Lydg.TG; etc.

   • Discrepancy is due to obsolete stencil.

     See below, under (6).

3. [OD col.] problems. All of these can be put off, but consider these for the future. We will NOT, for the moment, link any of the OD col quots. to the HB OD col. blanket entry.
   • Work is commonly quoted, but has no MED stencil.

     pfs will regularize the form so far as that can be done safely. and later consider creating a HB entry for it if there is sufficient information to do so.

   • Work is rarely quoted, but has no MED stencil.

     leave it alone in the MED. We will later consider creating a HB entry for it if there is sufficient information.

   • There is a special ODcol stencil given in HB.

     Send word to pfs so he can recast the citation in the MED so as to remove the page ref after the stencil. He will then map it to the HB stencil in a <grp>.

   • No special ODcol stencil given in HB, but one could be.

     Do nothing for the moment, but later consider adding HB stencil to existing HB entry, then recast MED citations to match.

4. Bracketed Cits problem.

   Stencil is used in MED only in bracketed citations, and is absent from HB.

   Though there are not very many of these, we will defer treatment of them and consider adding them later on the grounds that no bibliographic ref in the MED should go unexplained. Leave unmatched now.

   Stencil is used in MED in both bracketed and non-bracketed cits, but bracketed cits sometimes have dates not allowed to non-bracketed cits in HB.

   Match, add use note to HB indicating that the work is cited from other dates for non-ME quots.

5. Preference problems.
   • MED has stencil w/o MS id, and date of non-preferred MS

     check quot. to see if (a) really from that MS and (b) if MS is really non-preferred for that line. Change HB (if it is really preferred) or map (if it is really non-preferred but accidentally left off MS) as appropriate.

   • MED has stencil w/o MS id, and date of preferred MS as well as of other MSS

     Assume the data is right: map to HB preferred-ms stencil.

   • MED has stencil w/o MS id, and date app. uniquely belonging to preferred MS

     Assume the data is right: map to preferred MS HB-stencil.

   • MED has stencil with preferred MS id, but drops comp date

     check to see if really preferred for that passage, and if so, if omission of comp date is legit. If another MS is really preferred, change HB; otherwise map as usual.

   • MED stencils do not mention MS, HB offers multiple preferred MSS, MS can be discovered entirely or mostly by looking at page ref.

     do not map these, since they will have to be checked manually. Instead, create, if necessary, med stencils that include the MS info ("(LdMisc 471)") and map THOSE. Then send note to pfs including the new dummy stencils and explaining the situation; he will manually add MS info to the med stencils based on the page ref.

   • MED stencils do not mention MS, HB offers multiple preferred MSS, MS can *not* be ascertained by looking at reference.

     Check if this is really true. Perhaps MED card or book records long list of page refs that we simply neglected to include in favor of passim. Change HB, then do as above, including note to pfs.

     If it is true, and if rarely used, check book manually to see which ms was used for which quot.

     If it is true, and if commonly used, or cannot determine info from book, need to change HB entry to use double mss in abbrev, as with PConsc. Change map file to match change in HB.

6. Obsolete stencil problems
   • Large general question: should we try to revise the MED's use notes for them? Many of the coverage ranges ("in A-Se", etc.) in the use notes are guesswork; many are wrong. Our current MED files are largely accurate representations of the MED (though I *may* have accidentally replaced some obsolete dates while repairing stencils), and could be used to make the USE notes more accurate. E.g., *Let.Alex. is supposedly used "through Q"--but it also appears in R1 and R2.

     NO, save files at this pre-merged stage, as at earlier stages, and let MED use them if it likes to correct its own use notes.

     Particular problems:

   • USE records change of date without linking that change to a change of edition.

     replace the old date with the new one in the MED w/o changing the USE note (except by qualifying it: "[In A-L] of print MED...").

   • USE records change of date, linking that change to a change of edition.

     These should have a the older version tagged as a stencil within the USE note of the hb entry. If they are missing, add stencil tags, add the new stencil to the map file, and attach the relevant <med>'s.

   • USE records change from * to non-* or vice versa, without listing full stencil for obsolete version.

     change HB to tag stencil in USE note, add that stencil to the map file, and map.

   • SAME situation, except change from * to or from non-* coincides with change of date. This is like change-of-date linked to change-of-edition, except that it is (in general) less ambiguous, since identity of edition is hidden, presence or absence of * is manifest.

     change date in MED, and modify use note as above, when we can, to tag stencil in the note and to indicate that the obsolete date applies only to the print MED.

Other problems

Same passages referred to with multiple stencils (e.g. "in ..." and "doc.in...")

If there is a principled difference between the two stencils, or if one can be found or invented, retain both stencils, change misbehaving examples to correct stencil in MED, emend ref or use note to reflect distinction.

If no principle difference can be found, or if there is a wholesale mess (e.g. the "...in Sur.Soc.35" stencils), remove and add MED stencils as necessary to create a coherent situation, list all changes in the use notes to the appropriate entries ("in print MED also quoted with stencils ... and ... ").

Overlap between regular and OD-col stencils

See separate OD-col guidelines.

What to do with new entries

1. Create new entries as needed to accommodate "missing" stencils
2. Copy stencil from AE file to temporary text file of new stencils
3. When all the new stencils for a batch or set of batches have been added, treat the text file as if it were an HB batch itself, i.e. run first ex-sss.pl and then addmed.pl on the file to create a map file
4. Map the 'missing' stencils to the new stencils in the temporary map file.
5. Cut and paste the new <grp>s from the temporary map file to the permament map files in their appropriate alphabetical place.