srs3 derivation method

[funderburkjim]

In examining some of the TODO cases mentioned in #2 , cases like acCe@tya seemed to have an easy derivation, namely acCa+itya.

Why was this derivation not found by the srs2 method ?

The reason turns out to be that

• The srs2 method requires the first element (acCa) to be the parent of acce@tya.

• acce@tya is an H2, and the parent (the previous H1 entry) is, in the current digitization of MW, the word acCambawkArAya. Here is part of the hierarchy of entries:

  1   1866   acCa   acCa   
  1   1866.1   acCawASabda   acCawA-Sabda   
  1   1866.2   acCawAsaMGAta   acCawA-saMGAta   
  1   1866.3   acCandaska   a-cCandaska   
  1   1866.4   acCambawkArAya   a-cCambawkArAya   
  2   1867   acCAi   acCAi   
  numerous other H2 entries omitted
  2   1888   acCAsf   acCAsf   
  2   1889   acCAsyand   acCAsyand   
  2   1890   acCeta   acCe@ta   
  2   1891   acCetya   acCe@tya   
  

  Since the parent is not acCa, the srs2 method failed.

Note: Those 4 H1 headwords acCawASabda through acCambawkArAya are from the supplement to MW. It could be argued that they should be coded as H2. If this were done, then the srs2 analysis of acCeta and acCetya would succeed (with the same derivations, acca+ita and acCa+itya as currently provided by the srs3 analysis.

[funderburkjim]

Here is the updated summary of analysis2 with the srs3 method:

 22100  NTD init
   587 DONE +wsfx1
 85766 DONE cpd1
  1418 DONE cpd1a
  4379 DONE cpd3
  1095 DONE cpd4
  1791 DONE cpd5
  3295 DONE cpd_nan
 12546 DONE gender
  1188 DONE inflected
 42771 DONE noparts
  8666 DONE pfx1
  2583 DONE pfx2
  1576 DONE pfxderiv
 15048 DONE srs2
   983 DONE srs3
  6492 DONE wsfx
  7980 TODO init

983 of the cases are now analyzed by srs3. It was a pleasant surprise to see this large number of new derivations.

[gasyoun]

983 of the cases are now analyzed by srs3. It was a pleasant surprise to see this large number of new derivations.

even without coding the H1 headwords from supplements as H2?