Jasp:
> Next week I shall be talking to the LAGB (see
> http://www.linguistlist.org/issues/12/12-515.html ) on Word Grammar, in
> a special session on constructions.

I expect to be there, though it depends on how much teaching I can
manage to cancel in the morning.

> I'm attaching below my draft handout for the talk (diagrams are
> represented as well as I can in text; material in square brackets is not
> to appear on the handout), and in a second message the abstracts of the
> other participants for your information and comments. I really would
> welcome your comments, not least so I can know to what extent my views
> reflect those of the wider 'Word Grammar community'.

I found it a bit too elliptical to comment on intelligently, so here are
a couple of minor comments on incidental matters.

I do think, tho, that you articulate the main properties of WG -- the
cognitive network and the uncompartmentalized isa hierarchy.

> a. WG isa Dependency Grammar (DG):-
> words (and the relationships between them) are the only primitives of
> linguistic structure
> [this means, for sentence structures, simply that we do not recognise
> phrasal categories and it means, for lexical/grammatical structures that
> they too refer only to (classes of) words or dependencies]

As we're all aware, an endocentric/headed PSG can be translated into
a DG, and a DG can be translated into an endocentric PSG in which all
nonterminal nodes are maximal. There is no fundamental incompatibility
between WG and PSG; it's just that WG thinks that no decent case has
been made for nonterminal nodes, and they can therefore be considered
redundant. So I think the DGness of WG is more one of its most salient
properties rather than one of its most essential properties.

That said, there are pros and cons to nonterminal nodes. The cons, Dick
has pointed out in some old articles in _Linguistics_. The pros include
adjunct/complement ordering and handling what WG has to handle by
invoking semantic phrasing.

> c. Reasons for doubting these are all lexical properties of GIVE:
> productivity: ADVANCE, ALLOCATE, ALLOT, ASK, ASSIGN, AWARD, BARGE, BASH,
> BAT, BEQUEATH, BOUNCE, BRING, BUNT, BUS, CABLE, CARRY, CART, CATAPULT,
> CEDE, CHUCK, CITE, CONCEDE, DRAG, DRIVE, E-MAIL, EXTEND, FAX, FEED,
> FERRY, FLICK, FLING, FLIP, FLOAT, FLY, FORWARD, GIVE, GRANT, GUARANTEE,
> HAND, HAUL, HEAVE, HEFT, HIT, HOIST, HURL, ISSUE, KICK, KICK, LEASE,
> LEAVE, LEND, LOAN, LOB, LUG, MAIL, MODEM, NETMAIL, OFFER, OWE, PASS,
> PASS, PAY, PEDDLE, PHONE, PITCH, POSE, POST, PREACH, PROMISE, PULL,
> PUNT, PUSH, QUOTE, RADIO, READ, REFUND, RELAY, RELAY, RENDER, RENT,
> REPAY, ROLL, ROW, SATELLITE, SCHLEP, SELL, SEMAPHORE, SEND, SERVE, SHIP,
> SHOOT, SHOVE, SHOVE, SHOW, SHUTTLE, SIGN, SIGNAL, SLAM, SLAP, SLIDE,
> SLING, SLIP, SMUGGLE, SNEAK, TAKE, TEACH, TELECAST, TELEGRAPH,
> TELEPHONE, TELEX, TELL, THROW, TIP, TOSS, TOTE, TOW, TRADE, TRUCK, TUG,
> VOTE, WHEEL, WILL, WIRE, WIRE (MONEY), WIRELESS, WRITE, YIELD. (Levin
> 1993:xx)
> ARRANGE, ASSEMBLE, BAKE, BAKE, BLEND, BLOW, BOIL, BOOK, BREW, BUILD,
> BUY, CALL, CARVE, CASH, CAST, CATCH, CHARTER, CHISEL, CHOOSE, CHURN,
> CLEAN, CLEAR, COMPILE, COOK, COOK, CROCHET, CUT, DANCE, DESIGN, DEVELOP,
> DIG, DRAW, EARN, EMBROIDER, FASHION, FETCH, FIND, FIX, FOLD, FORGE, FRY,
> GAIN, GATHER, GET, GRILL, GRIND, GROW, HACK, HAMMER, HARDBOIL, HATCH,
> HIRE, HUM, IRON, KEEP, KNIT, LEASE, LEAVE, LIGHT, MAKE, MINT, MIX, MOLD,
> ORDER, PAINT, PHONE, PICK, PLAY, PLUCK, POACH, POUND, POUR, PREPARE,
> PROCURE, PULL, REACH, RECITE, RENT, RESERVE, ROAST, ROLL, ROLL, RUN,
> SAVE, SCRAMBLE, SCULPT, SECURE, SET, SEW, SHAPE, SHOOT, SING, SLAUGHTER,
> SOFTBOIL, SPIN, SPIN, STEAL, STITCH, TOAST, TOSS, VOTE, WASH, WEAVE,
> WHISTLE, WHITTLE, WIN, WRITE. (Levin 1993:xx)
>
> creativity: POST, FAX, EMAIL, TEXT, SMS; Augusta waggled me her bid with
> her ears.

I'm not sure exactly what your point is, but IMO the facts are that
there is an open class of ditransitives, such than any transitive verb
can be ditransitive, in which case it will be construed as having
one of a certain range of senses, and that this still leaves some
ditransitives unaccounted for (because they don't have one of this
certain range of senses), and therefore must be lexically specified
as ditrans. GIVE is an example of this latter type, because plenty of
plesionyms of GIVE can't be ditrans.

--And.