\documentclass[12pt]{article}
%\usepackage{phonetic-latex2e,tree-dvips,lingmacros,formalstuff}
\usepackage{tree-dvips,lingmacros,formalstuff}
\usepackage{array,pstricks,colortab, pifont, fancybox, rotating, graphicx}
\usepackage{times}
\usepackage[safe]{tipa}


\newcommand{\sent}{\enumsentence}
\newcommand{\sents}{\eenumsentence}
\newcommand{\R}{\ding{43}}
\newcommand{\otinput}{\textsc{input}}
\newcommand{\otoutput}{\textsc{output}}
\newcommand{\cand}{\textsc{cand}}
\newcommand{\otgen}{\textsc{gen}}
\newcommand{\eval}{\textsc{eval}}
\newcommand{\otlfg}{\textsc{ot-lfg}}
\newcommand{\moreharmonic}{$\succ$}
\newcommand{\morharm}{$\succ$}
\newcommand{\outranks}{$\gg$}
\newcommand{\struct}{{\sc *Struct}}
%\newcommand{\faith}{{\sc Distinct(R,T)}}
\newcommand{\faith}{{\sc Faith(Rec)}}
\newcommand{\nppro}{{\sc *NP Pron}}
\newcommand{\propro}{{\sc *XP Pron}}
\newcommand{\harmlocal}{{\sc Harmony(1,2)}}
\newcommand{\harm}{{\sc Harmony}}
\newcommand{\harmpron}{{\sc Harmony(Pron)}}
\newcommand{\OO}{{\sc OO-Primacy}}
\newcommand{\pro}{\textsc{pro}}
\renewcommand{\k}{\textpolhook}
\newcommand{\rsub}[1]{$_{\mbox{\it{#1}}}$}
% take out the following for seminar slides
\renewcommand{\small}{}
\renewcommand{\red}{}
\newcommand{\faithother}{{\sc Faith\sub{other}(Rec)}}
\newcommand{\faithgive}{{\sc Faith\sub{give}(Rec)}}
\newcommand{\faithlatin}{{\sc Faith\sub{latinate}(Rec)}}

\begin{document}


\begin{center}
{\Huge On the Gradience of the Dative Alternation \par}

\bigskip\bigskip\bigskip


{ {\sc\large Joan Bresnan and Tatiana Nikitina}}\footnote{We thank
    Beth Levin, Chris Manning, Lena Maslova, K.P.\ Mohanan, Tara
    Mohanan, Ivan Sag, Peter Sells, Tom Wasow, Annie Zaenen, and the
    audience at an MIT Linguistics Department Colloquium on May 2,
    2003, for comments on various aspects of this work.  We alone are
    responsble for the contents.  This work is based in part on
    research supported by the National Science Foundation under Grant
    No.\ BCS-9818077.  Some diagrams were prepared with Boersma and
    Weenink's (2000) freely available Praat program.}\bigskip

Stanford
    University\\ Draft of May 6, 2003\\[2ex]\bigskip\bigskip


\noindent {\sc Abstract}\end{center}

\begin{quote}
The present study addresses the gradience of the dative alternation.
It is shown that central evidential paradigms that have been used to
support both lexical semantic (Pinker 1989, Krifka 2001) and
constructional (Goldberg 1995, Kay 1996, ao) explanations for the
choice of dative constructions are not well founded empirically.  Some
widely repeated reports of intuitive contrasts in grammaticality
appear to rest instead on judgments of pragmatic probabilities.  An
informational theory of the dative alternation is supported by the
results of a corpus study on the distribution of person across dative
NP and PP recipients in spoken English, and a formal model of the
theory is given within the framework of stochastic Optimality Theory.
The model incorporates lexical variation and leads to the expectation
that the same pattern of person/argument alignment that appears
gradiently in the English dative alternation should appear
(near-)categorically in other languages, which is borne out by the
Nilo-Saharan language Kanuri.
\end{quote}\bigskip


     What drives the dative alternation?  Two broad classes of
approaches have been taken to answering this question: the semantic
and the informational.  

   Semantic approaches to explaining the dative alternation have
attempted to map each of a number of fine-grained semantic classes of
dative verbs and idioms onto a unique syntax (Green 1974; Oehrle 1976;
Gropen, Pinker, Hollander, Goldberg, and Wilson 1989; Pinker 1989;
Speas 1990; Levin 1993; Marantz 1993; Goldberg 1995; Pesetsky 1995;
Davis 1997; Harley 1996, in press; Arad 1998; Kay 1996; Bruening 2001;
Krifka 2001, ao).  We show in the first two parts of this study that
central evidential paradigms that have been used to support both
lexical semantic (Pinker 1989, Levin 1993, Krifka 2001) and
constructional (Goldberg 1995, Kay 1996) explanations for the choice
of dative constructions are not well founded empirically.  Some widely
repeated reports of intuitive contrasts in grammaticality appear to
rest instead on judgments of pragmatic probabilities.

    Informational approaches attribute the use of alternative dative
syntax to contextual or processing factors such as information
structure, animacy, definiteness, and end weight (Halliday 1970;
Smyth, Prideaux, and Hogan 1979, Erteshik-Shir 1979, Ransom 1977, 1979,
Giv\`on 1984, Thompson 1990, 1995; Hawkins 1994, Collins 1995; Davidse
1996a,b; Arnold et al 2000; Snyder 2001, Levin and Rappaport Hovav
2002).  In the third part of this study we provide an
informational model of the dative alternation using the framework of
stochastic OT.  We present results from a corpus study showing that
there is a harmonic alignment of person with the syntactic argument
type of the dative recipient in the parsed {\sc Switchboard} corpus of
spoken English (Godfrey, Holliman, and McDaniel 1992; Marcus,
Santorini, and Marcinkiewicz 1998).  The person/argument alignment is
independent of previously studied effects of length and syntactic
complexity (Hawkins 1994, Wasow 2002), pronominality and definiteness
(Giv\'on 1984; Thompson 1990, 1995; Collins 1995).  Our stochastic OT
model of person/argument alignment captures the gradience of this
phenomenon as well as the divergences in the probability distributions
of different classes of verbs.  The model leads to the expectation
that near-categorical person and pronominality splits in the same
direction as the quantitative alignment in English should exist in
other languages.  In the fourth part of this study we show that this
expectation is borne out in the Nilo-Saharan language Kanuri
(Hutchison 1981).


\section{Is there a dative alternation?}

   As mentioned above, semantic approaches to explaining the dative
alternation map each of a number of fine-grained semantic classes of
dative verbs and idioms onto a unique syntax.  Abstracting away from
differences in choice of syntactic representation, one general idea
that has appeared in a number of these approaches is that dative
verbs or idioms which have possessive semantics as in (i) are uniquely
associated with the dative NP syntax \mbox{[V NP NP]}, while datives
with allative semantics as in (ii) are uniquely associated with the
dative PP syntax \mbox{[V NP [to NP]\pslabel{pp}]}:\footnote{On the
possessive meaning of the dative NP construction see also Herslund
(1986) and Davidse (1996b).}
\medskip

\begin{tabular}{rlccr}
(i) & `x causes y to have z' &(possessive)&  $\Rightarrow$ &NP V NP 
NP  \\
(ii) & `x causes z to go to/be at y' &(allative)&  $\Rightarrow$& NP V NP
[to NP]   \\
\end{tabular}\medskip

\noindent On these approaches, the dative NP and PP constructions are
not alternative expressions of the same meaning, they are expressions
of different meanings.  Hence, on this view there is no true dative
alternation.

     Arguments in favor of this approach list a number of semantic
restrictions on the dative alternation, such as those in
\exno{1}--\exno{3}:

\sents{
\item I threw the box to John.  $\sim$ I threw John the box.
\item I lowered the box to John. $\not\sim$ *I lowered John the box. 
\item[](Pinker 1989: 110--111; Levin 1993: 46, 114)
 }

\sents{
\item Ann faxed the news to Beth. $\sim$  Ann faxed Beth the news
\item  Ann yelled the news to Beth . $\not\sim$   *Ann yelled Beth the news.
\item[](Krifka 2001)
 }
\sent{
 The lighting here gives me a headache. $\not\sim$
 *The lighting here gives a headache to me.

(Marantz 1993; Bruening 2001: 261)
 }

\noindent For example, giving someone a headache is causing them to
have a headache, not transferring the headache from one location to
another.  Hence, by virtue of its meaning, it is argued, this idiom
occurs only in the possessive dative NP construction (i) and does not
alternate.  Likewise, the meaning of `throwing' specifies the causing
event in the schema (i), while the meaning of `lowering' specifies
both the causing event and the movement event in schema (ii), since
there is a homomorphoric mapping between the two events in lowering
actions (Krifka 2001, Pinker 1989).  Therefore `lowering' and
similar verbs cannot have the syntax associated with schema (i)
because it omits a essential part of their meaning.  With yelling in
\exlet{-1}{b}, there is ``a homomorphism between speech production
(e.g.\ the activity of yelling) and the transfer of information,''
according to Krifka (2001), while with faxing there is no homomorphism
between the causing event and the movement event; only the initial
stage of the transfer is specified as with {\it throw} in
\exlet{-2}{a}.

     When the same verb appears with both dative NP and dative PP
syntax on this account, the meanings of the two constructions differ.
Either the verbs \textit{throw, fax}, and the like are lexically
polysemous, or polysemy is imposed by the differing constructional
contexts they appear in, depending on the specific grammatical assumptions
(lexical or constructional) of the approach.

     A challenge for the approach is the fact that alternating dative syntax
can be found in contexts of repetition, as in the following attested
examples illustrate.

\sent{
 ``You don't know how difficult it is to find something which
will please everybody---especially the men.''

   ``Why not just {\bf give them cheques}?' I asked.

   ``You can't {\bf give cheques to people}.  It would be
   insulting.''\footnote{Davidse (1996a: 291), from Graham
   Green (1980) \textit{ 
    Doctor Fischer of Geneva or the Bomb Party}.  London: The Bodley
  Head)} 
}
\sent{
 ``You {\bf carrying a doughnut to your aunt} again this morning?''
   J.C. sneered.  Shelton nodded and turned his attention to a tiny TV
   where ``Hawaii Five-O'' flickered out into the darkness of the little
   booth.  ``Looks like you {\bf carry her some breakfast} every
   morning.''\footnote{www.flagpole.com/Issues/12.23.98/shortstory.html}
}

\noindent Krifka (2001), building on Gropen et al.\ (1989) and Pinker
(1989),  proposes for {\it give} that ``every transfer of
possession entails an abstract movement event in the dimension of
possession spaces.''\footnote{The latinate verbs (\textit{donate, contribute},
etc.) remain an exception to this generalization for
morphophonological reasons.}$^{,}$\footnote{The {\it give a headache}
idiom is not affected by this meaning postulate, according to Krifka (2001),
because the theme does not just change possession but comes into
existence.}  This proposal makes some polysemies empirically
indistinguishable from the monosemy hypothesis (Levin and Rappaport
Hovav 2002), which asserts that when verbs in the broad semantic
classes (i) and (ii) have recipient rather than purely spatial
arguments, each can occur with both dative NP and PP syntax.

The question then becomes, What is the scope of these equivalenced
polysemies or monosemies?  We have found that it is far greater than
has previously been recognized.

\subsection*{Case 1: `verbs of imparting of force'}

``Verbs of instantaneous imparting of force in
some manner causing ballistic motion'' occur with
both dative NP and PP syntax:

\sent{
 Lafleur throws\slash tosses\slash flips\slash slaps\slash
  kicks\slash pokes\slash flings\slash blasts him the
  puck; he shoots, he scores!

(cf.\ Lafleur throws\slash tosses\slash flips\slash slaps\slash
  kicks\slash pokes\slash flings\slash blasts the
  puck to him; he shoots, he scores!)}

\noindent In contrast, according to Pinker (1989: 110--111) and Krifka
(2001) ao, ``verbs of continuous imparting of force in some manner
causing accompanied motion'' occur only in the dative PP construction:

\sent{\bstar I carried\slash pulled\slash pushed\slash
  schlepped\slash lifted\slash lowered\slash hauled John the box.

(cf.\ I carried\slash pulled\slash pushed\slash
  schlepped\slash lifted\slash lowered\slash hauled the box to John.)}

\noindent Of verbs of continuous imparting of force (\textit{carry, pull, push,
schlep}), Pinker (1989: 103) writes:

\begin{quote}
   Though they are \textit{cognitively} construable as resulting in a
   change of possession (if the object is pushed over to a person with
   the intent of giving it to him), they are not
   \textit{linguistically} construable as such because the licensing
   linguistic rule is not stated broadly enough to apply to them.
\end{quote}

     Yet we find from an examination of WWW documents that
verbs of continuous imparting of force are linguistically construable
as depicting changes of possession, and are in current use.  The
following examples are a selection of our findings.

{\sent{\textsc{verbs of continuous imparting of force}
\label{carry-verbs} }

\begin{quote}

   Karen spoke with Gretchen about the procedure for registering a
   complaint, and {\bf hand-carried her a form}, but Gretchen never
   completed it. \\ {\small 2 June 1999, Nampa Controversy Summary ---
   Idaho Library Association}, 
   {\small www.idaholibraries.org/nampa.controversy.summary.htm}\medskip

   As Player A {\bf pushed him the chips}, all hell broke loose at the
   table.\\ {\small
   www.cardplayer.com/?sec=afeature\&art\_id=165}\medskip

   Therefore, when he got to purgatory, Buddha {\bf lowered him the
   silver thread of a spider} as his last chance for salvation.\\
   {\small www.inch.com/~fujimura/ImofGrmain.htm}\medskip

   Nothing like heart burn food.  ``I have the tums.'' Nick joked.  He
   {\bf pulled himself a steaming piece of the pie}. ``Thanks for
   being here.''\\ {\small
   www.realityfanfiction.addr.com/storm3.html}\medskip

   ``Well\ldots it started like this\ldots'' Shinbo explained while
   Sumomo {\bf dragged him a can of beer} and opened it for him, ``We
   were having dinner together and\ldots'' \\ {\small
   www.angelfire.com/wa2/bozyby/hold1.html}

\end{quote}
}
\paragraph{Notes.}   
The context of the second example is a tournament poker game:
the card players are seated together at a card table and have bet
varying amounts by putting poker chips into the pot; this is done by
placing some of one's own poker chips onto a common area of the table
for the chips of betting players.  Whoever wins the pot receives all
of the chips, which can be pushed across the table to the winner.

The last example is from a Chobits fanfiction piece: Shinbo is
sitting on a tatami mat with his interlocutor (Hideki). Sumomo is a
very small servant robot, small enough to dance on a table, climb up
his master's leg and perch on his shoulder.  Sumomo serves the beer to
the visitor Shinbo by dragging a can to him.


\subsection*{Case 2: Verbs of communication}

     Another widely repeated contrast occurs among verbs that can be
used for describing types of communication (Pinker 1989, Levin 1993,
Krifka 2001 ao).   ``Verbs of instrument of communication''
have uses with both dative NP and PP syntax:

\sent{ Susan cabled\slash emailed\slash faxed\slash phoned\slash
     telegraphed\slash \ldots\ Rachel the news.

(cf.\ Susan cabled\slash emailed\slash faxed\slash phoned\slash
     telegraphed\slash \ldots\  the news to Rachel.)}

\noindent In contrast, ``verbs of manner of speaking'' are marked as
ungrammatical with dative NP syntax:

\sent{\bstar Susan whispered\slash yelled\slash mumbled\slash
      barked\slash muttered\ldots\ Rachel 
      the news.

(cf.\ Susan whispered\slash yelled\slash mumbled\slash
      barked\slash muttered\ldots\ 
      the news to Rachel.)}


Despite the reported ungrammaticality of verbs of manner of speaking
with dative NP syntax, we again find representatives of the starred
types of examples in current use:

\sent{\textsc{manner of speaking verbs}\label{yell-verbs}\medskip

%\begin{quote}

   Shooting the Urasian a surprised look, she {\bf muttered him a
   hurried apology} as well before skirting down the hall. \\ {\small
   www.geocities.com\slash cassiopeia\_sc\slash fanfiction\slash
   findthemselves.html} \medskip

   ``Hi baby.''  Wade says as he stretches. You just {\bf mumble him
   an answer}.  You were comfy on that soft leather couch. Besides
   \ldots\\ {\small
   www.nsyncbitches.com/thunder\slash fic/break.htm}\medskip

   The shepherd-dogs, guardians of the flocks, {\bf barked him a
   welcome}, and the sheep bleated and the lambs pattered round him.\\
   {\small www.litrix.com\slash raintr\slash raint009.htm}\medskip

   I think he was poking fun at the charges that Blackmore has been
   making that he chronically forgets words --- he went over to Jon Lord
   during `Smoke' and seemed to be getting Jon to {\bf yell him the
   words}!!\\ {\small
   www.thehighwaystar.com\slash reviews\slash namerica\slash
   asbuandr.htm}\medskip 

   I still can't forget their mockery and laughter when they heard my
   question. Finally a kind few (three to be exact) came forward and
   {\bf whispered me the answer}.\\ {\small
   www.bangla2000.com\slash mboard\slash vbulletin.asp?ID=1462}

%\end{quote}
 }

\subsection*{Case 3: \textit{give} NP NP idioms}

Idioms have been long and widely cited as showing that the dative NP
and dative PP constructions differ semantically.  To quote just one of
many authors who have repeated this claim, Davis (1997: 41) writes
of idioms like {\it give me a headache} and {\it give him a
punch}:
 

\begin{quote}
   These sentence[s] denote situations in which a participant
   acquires a headache or receives a punch, but the headache and the
   punch cannot be said to be transferred from one location to
   another.  Accordingly, the ditransitive one is the only appropriate
   one in these instances.
\end{quote}


\noindent  Yet these idioms are in fact used with dative PP syntax, as are all
possibly idiomatic {\it give} NP NP collocations we found in the {\sc
Switchboard} corpus.  The following is a  representative selection.

\sent{\textsc{give a headache to} \label{headache}\medskip

%\begin{quote}

  sending a copy to every elector is a nice gesture, but futile,
  because it is unreadable, guaranteed to {\bf give a headache to
  anyone who looks hard at the small print}.\\ {\small (The Guardian
  (London), September 17, 1992, p. 23; Nexis) [from Levin and
  Rappaport Hovav 2002]}\medskip

  From the heads, offal and the accumulation of fishy, slimy matter, a
  stench or smell is diffused over the ship that would {\bf give a
  headache to the most athletic constitution}.\\ {\small
  www.downhomer.com\slash Webmag\slash 2000\slash 0007\slash
  page36.html}\medskip

  She found it hard to look at the Sage's form for long. The spells
  that protected her identity also {\bf gave a headache to anyone
  trying to determine even her size}, the constant bulging and
  rippling of her form gaze Sarah vertigo.\\ {\small
  http:\slash\slash lair.echidnoyle.org\slash rpg\slash log\slash
  27.html}\medskip 

  Design? Well, unless you take pride in {\bf giving a headache to
  your visitors} with a flashing background?no.\\
  http:\slash\slash
  members.tripod.com\slash$\sim$SailorMoonWorstOfWeb\slash
  archive\slash Run\-Jan01.\-html

%\end{quote}
 }

\sent{\textsc{give a punch to} \label{punch}\medskip

%\begin{quote}

  When the corpse was bloodless, he got up and grinned to
  Ethan-vampire, oh so happy.  ``Oh yesssss!'' He {\bf gave a punch to
  his old mate}. ``Let's find a bar, Ethan.'' \ldots\\
  vampirecows.com\slash odd\slash authors\slash anne\slash
  draculaannebg01.html\medskip

  ``Well, mate, you asked for it.''- And he {\bf gave a punch to the guy}
  in the middle of his face, splotching \ldots\\
  www.fortunecity.com\slash tattooine\slash tolkien\slash 176\slash
  tekrats.htm\medskip

  All three headed toward Mulan. She dropped kicked the first. Next
  she {\bf gave a punch to the second man}. He blocked so she grabbed his
  arm and flipped him. \ldots\ \\ members.tripod.com\slash Xi\_Xiao\slash
  family002.html \medskip

  She {\bf gave a punch to the evil reporter that had asked the dumb ass
  question}. \\
  http:\slash\slash pub56.ezboard.com\slash
  ffoxprintsfrm5.showMessage?topicID=10.topic

%\end{quote}
 }
\sent{{\textsc{give a break to}}\label{break}\medskip

%\begin{quote}
  PUC {\bf gives a break to big users of energy}.\\
  www.sacbee.com\slash content\slash politics\slash story\slash
  5114554p\-6120694c.html\medskip

  ``Why can't we {\bf give a break to the people who organise them} [the
  matches]?''\\  www.rediff.com\slash cricket\slash 2002\slash mar\slash
  22kapil.htm\medskip

  {\bf Give a break to the overburdened who have no place to rest}.\\
  www.csmonitor.com\slash durable\slash 1999\slash 11\slash 03\slash
  p15s1.htm\medskip 

  That's been the fairest way I can think of to protect the people who
  do register, and still {\bf give a break to the people who have
  contributed to the project}\ldots\\
  www.qflux.net\slash wwwboard\slash messages\slash 1057.html\medskip

  They wonder what citizenship means if you {\bf give a break to people who
  are here illegally}.\\  www.usbc.org\slash profiles\slash
  0202citizenshipmatter.htm \medskip

%\end{quote}
 }
\sent{{\textsc{give a hard time to}}\label{hard}\medskip

%\begin{quote}

  The silly clowns sometimes {\bf give a hard time to the emperor}.\\
  www.math.ohio-state.edu\slash $\sim$econrad\slash lang\slash ln.html\medskip

  The Necromancer has a wide area of spells he can use to either stay
  out of trouble or {\bf give a hard time to his opponents}.\\
  www.ultimategamers.com\slash diablo\slash necromancer\slash
  necromancer.html\medskip

  Those who have come before traditionally {\bf give a hard time to those
  who have just come}\\ www.mcny.org\slash byron\slash GCAintro.htm\medskip
%\end{quote}
}
\sent{{\textsc{give grief to}}\label{grief}\medskip

%\begin{quote}
  Still, I took it back today and {\bf gave some grief to the assistant} and
  came out with a better scanner than I had paid for on Tuesday\\
  scribblepad.co.uk\slash archive\slash april2002.html\medskip

  He {\bf gave grief to those taking their time near the rear}, I remember
  watching him from outside the bus while we stood on the yellow
  footprints. \\ pages.sbcglobal.net\slash e8usmcdi\slash 1stday.html

%\end{quote}
 }

\noindent For further discussion of idioms in relation to usage data, see Snyder
(2001), Davidse (1998), and especially  Levin and Rappaport Hovav
(2002).

\subsection*{Case 4: Verbs of prevention of possession}

     Even the verbs {\it cost} and {\it deny}, which are widely
described as occurring only with dative NP syntax, we found to
alternate.  Contrast \exlet{1}{a,b} (Krifka 2001, among many others)
with \exno{2}:

\sents{
\item The car cost Beth \$5000. $\not\sim$ *The car cost \$5000 to Beth.
\item Ann denied Beth the ice cream. $\not\sim$ *Ann dened the ice
  cream to Beth. 
 }

\sent{\textsc{cost \ldots\ to}\medskip

%\begin{quote}

  The IRS is unionized, and the union apparently has the fear that
  outsourcing will {\bf cost jobs to their members}.\\
  www.collectionindustry.com\slash agencyNews\slash
  feedback.cfm?issue=4\medskip 

  Any reduced rate, however, will still {\bf cost jobs to Californians in
  the teleservices profession}, drive up costs, increase inefficiency,
  and place an undue restraint on technology.\\ {\sc fight against proposed
  predictive dialer ban in california continues},
  http:\slash\slash www.ataconnect.org\slash htdocs\slash
  govtrel\slash news\slash 2000\slash aug/08-18\slash
  ca\_ab2721update.htm\medskip  

  He did so thinking it would {\bf cost nothing to the government}.\\
  www.stuttgart.army.mil\slash community\slash Citizen\slash
  2000/0926\slash cheapcall.htm\medskip 
 
}
%\end{quote}

\sent{{\sc deny \ldots\ to}\medskip

Most grievances will involve only a dispute between the grievor and
the employer. The employer has underpaid, or disciplined, or {\bf denied a
leave to a teacher}; resolution of the grievance does not impact
directly on others. \\
www.bctf.ca\slash bargain\slash grievances\slash backgrounder.html\medskip

definition of 'abnegate'. {\it The American Heritage Dictionary of the
 English Language}, 4th Edition:

1.\ To give up (rights or a claim, for example); renounce. 2.\ {\bf To deny
(something) to oneself}: The minister abnegated the luxuries of life.

www.bartleby.com\slash 61\slash 83\slash A0018300.html\medskip

After all, who could {\bf deny something to someone so dedicated
to the causes of international friendship and collaboration}?\\
www.eawc.org\slash 7forum\slash loula\_greece.html\medskip

%\end{quote}

 }

\subsection*{The answer to the question: Yes}

     We observe that {\it give a headache, give a punch, give a break,
give a hard time, and give grief} have fixed meanings which are
self-evidently constant across the dative NP and dative PP contexts.
Likewise, the verbs of deprivation of possession {\it cost} and {\it
deny} mean the same in the dative PP constructions.  The verbs of
continuous imparting of force {\it drag, carry, push, pull}, and {\it
lower} still specify the same distinguishing manners of motion in
the dative NP contexts cited as they do in the dative PP context.
Likewise, the manner of speaking verbs {\it mutter, mumble, bark,
yell, whisper} continue to specify the same characteristic emissions
of sound continuously accompanying the speech acts in the dative NP
contexts cited as in the dative PP contexts.

    The imagined inability of these verbs, idioms, and constructions
to be used in one of the two dative constructions has provided
central evidential paradigms for the idea that
differing semantics dictates the differing syntactic expressions of
the dative.\footnote{Reported entailments of completion or
affectedness accompanying the dative NP construction have been shown
to be defeasible implicatures (Levin and Rappaport Hovav 2002 and
references).} On closer inspection, there seems to be no reason at all
to reject the dative alternation for a much wider range of verbs and
idioms than previously recognized.

     We conclude that the dative alternation exists, whether it is
analyzed in terms of alternative syntactic expressions of the same
meaning (Levin and Rappaport Hovav's 2002 monosemy hypothesis) or of
meaning postulates that create equivalences in truth conditions across
semantically differing dative NP and dative PP constructions (like
Krifka's 2001 equivalenced polysemies extended to a much wider range
of uses of verbs).

     Note that none of the starred example types are found in the parsed
{\sc Switchboard} corpus,\footnote{The {\sc Switchboard} corpus is a
database of spontaneous telephone conversations spoken by over 500
American English speakers, both male and female, from a great variety
of speech communities (Godfrey et al.\ 1992).  The conversations
average 6 minutes in length, collectively amounting to 3 million
words.  We used the parsed portion of this corpus (released as part of
the Penn Treebank, Marcus et al.\ 1993), which contains 1 million
words.} but all of them occur in the much larger corpus of web
documents.\footnote{The WWW is estimated to contain 47,000,000,000
words of English as of February 2000 (Baayen 2003, citing Grefenstette
2000).}  They appear not to be grammatically impossible, but just improbable.


\section{A systematic bias in grammaticality judgments?}

The usage data for the dative alternation raise an interesting
problem.  Our own linguistic intuitions agree with those of the
linguists cited (Pinker, Krifka, Levin, et al.), in that we perceive
the contrasts in the constructed examples.  At the same time we judge
the web examples given above to be grammatically
possible.\footnote{Thus we do not classify our usage data given above with the
sporadic adult errors of the types recorded by Gropen et al.\ (1989:
251).}  What then do our intuitions correspond to?

     Although both dative NP and dative PP constructions can be used
to express transfers of possession, as our examples show (see
also Levin and Rappaport Hovav 2002), the fact is that there is
a strong skewing of the syntax of alternating dative verbs toward the dative NP
construction in conversational English usage, as measured by the
parsed Switchboard corpus of English telephone conversations.\footnote{
The parsed {\sc switchboard} corpus is a database of
spontaneous telephone conversations spoken by over 500 American English
speakers, both male and female, from a great variety of speech communities
(Godfrey et al.\ 1992).  The conversations average 6 minutes in length,
collectively amounting to 3 million words.  We have used the parsed portion of
this corpus  (released as part of the Penn Treebank, Marcus et al.\ 1993),
which contains 1 million words.\label{swbd}}  See
Table \ref{allVerbs}.\footnote{This count excludes nonalternating
 uses of dative verbs.  See n.\ \ref{tell-you-what}.}  
\begin{table}[htbp]\caption{Alternating dative verbs in the parsed Switchboard
    corpus} \label{allVerbs}\bigskip

\hfil\begin{tabular}{lrrrr}
                &   NP NP  &    NP PP\rlap{\sub{to}} &  total:\\\hline
   {\it give}  &    226   &      35       &  261 &  {\it give}: NP NP = 87\%\\
   other verbs  &    291   &      69       &  360 &   other v's: NP NP = 81\%\\\hline 
    total:      &    517   &      104       &  621\\
                &          &               &   {\it give} = 42\rlap{\%}  \\
                &          &               &   of all instances\\
                &          &               &   of dative verbs\\
\end{tabular}
\end{table}
If we take {\it give} to be prototypical of the class of transfer of
possession verbs, then dative NP syntax is by far the preferred syntactic
expression for this class of verbs.

     Now transfers of possession may occur in many ways.  In sports
like hockey, possession of the puck can take place by means of a
number of sudden actions in play, and there is much varied discourse
about it.  In the world more generally, or at least in present-day
American life, if a person accompanies and holds, clings to, or
otherwise stays in contact with a possession, it seems to us less
likely that a transfer of possession is going on, and in many cases
there is probably much less talk about it than about possession of the
ball or puck or whatever in sports.  Carrying people things as a
transfer of possession is surely more common in situations where
walking is a major mode of transportation.  The previously given web
examples are from present-day English, but many examples of {\it carry}
with dative NP can be found on the web in depictions of life in rural
areas, often predating the rise of the automobile.

\sent{Pre-automotive uses of ditransitive {\it carry}\medskip

   Aurie and Pearl went to Humboldt that afternoon. I went back to
   Mrs. Kate's to {\bf carry her some mustard salad}.\\ {\small
   (www.rootsweb.com\slash $\sim$tngibson\slash Bios\slash
   mayfield1894.htm\\{} [from Fidelia  Mayfield Diary 1892])}\medskip

   ``This evening she was late starting dinner because her second
   granddaughter has a cold, and she had to {\bf carry her some pepper
   sauce for her cough}.''\\ {\small (www.fictionwise.com\slash
   ebooks\slash eBook842.htm\\{} [from a novel set in the Civil War
   period])}\medskip

   Polly had been sick and Sara wanted to {\bf carry her some food}.\\
   {\small (www.lrwma.com\slash happenings\slash {\sc
   happenings1}.htm\\{} [from Happenings around Leatherwood Mountain
   in the Early 1900s])}\medskip

   ``Go, my dear, and see how thy grandmamma does, for I hear she has
   been very ill; {\bf carry her a custard and this little pot of
   butter}.''\\ {\small (www.azstarnet.com\slash reading\slash
   reading22.html\\{} [Little Red Riding Hood])}

 }


\noindent For the same reasons, pushing is probably less likely to be
discussed as a mode of transferring possession than carrying, with
pulling perhaps less so, and lowering and dragging the least.  These
observations raise the possibility that our grammaticality judgments
of the contrasting pairs of examples are being systematically biased
by the probability of similar descriptions of the event types depicted
by the examples.   

    In summary, our hypothesis is this.  We can use both dative NP and
dative PP syntax to express transfers of possession, but the
prototypical uses of giving are heavily biased toward the dative NP
construction.  Now transfers of possession are more likely to be described
in the discourse of sports where motional verbs of instantaneous
imparting of force ({\it throw, toss, kick, flip, slap, fling,
\ldots}) are heavily used than in discourse about dragging, lowering,
pushing, pulling, and even carrying these days.  Hence, we are more
likely to judge verbs in the {\it throw} class as acceptable with
dative NP syntax than verbs in the {\it drag} class.

    We can pursue a similar line of thinking about the verbs of
communication.  Both types of communication verbs (`verbs of means of
communication' and `verbs of manner of speaking') are grammatically
possible with alternative dative syntax, yet with the first type
dative NP syntax seems to be preferred in grammaticality
judgments.  Again we may ask, what do these intuitions correspond to?

     Notice that activities of cabling, emailing, faxing, phoning,
telegraphing, and the like almost always involve communication---that
is transfers of the possession of information.  The most frequent verb
of communication that occurs in dative constructions is {\it tell}.
Over 99\% of all dative uses of {\it tell} in the parsed Switchboard
corpus occur in the dative NP construction.\footnote{This count
excludes nonalternating uses such as concealed questions ({\it I will
tell you another plant that is purply}) and 17 occurrences of the
fixed expression {\it I('ll) tell you what}. \label{tell-you-what}}

    In contrast to the activities described by the manner of
communication verbs, whispering, yelling, mumbling, barking,
muttering, and the like are more often, to varying degrees,
noncommunicative.  When used intransitively and with certain
directional phrases, the manner-of-speaking verbs ``describe the
physical characteristics of a sound'' rather than ``an intended act of
communication by speech'' (Zwicky 1971: 225, 226):

\sents{
\item   He whispered\slash yelled\slash mumbled\slash barked\slash
   muttered (but he wasn't saying  anything).
\item   He whispered\slash yelled\slash mumbled\slash barked\slash
   muttered at us\slash in our direction.
 }

\noindent In fact, a tgrep query of the Switchboard corpus yields 17
occurrences of these five manner of speaking verbs, of which 12 are
noncommunicative, 3 are semi-communicative (like ``yelling for help'',
which may not successfully communicate because an interlocutor or even
an audience is not necessarily present), and only 2 have complements
which denote ``the products of a speech act''.

     Granted that the uses of manner of speaking verbs are probably
disproportionately describing noncommunicative activities, why should
their communicative uses favor the dative PP over the dative NP?
Zwicky (1971: 226) observes that the directional {\it at, toward}
phrases that modify manner of speaking verbs are in complementary
distribution with the {\it to} PPs.  

\sent{
   He whispered/yelled/mumbled/barked/muttered at us\slash in our
   direction (*to John).}

\noindent This fact suggests that these verbs have a variant of the
allative type lexical semantics; here the PP denotes the orientation
of the actor toward the goal rather than a path of movement.  With
these verbs the theme argument is usually a noncommunicative sound and
less often the product of a speech act.  The same PP syntax expresses
both situations. 

     These observations are only suggestive, but they motivate our
conjecture that {\it grammaticality judgments of contrasting pairs of
examples may be systematically biased by the probability of similar
descriptions of the event types depicted by the examples.}  

     Note that it is the probability of the \emph{descriptions} of
event types, not the events themselves, that we conjecture to be
important in judging grammaticality.  We have no idea whether yelling
or muttering events are more or less probable than emailing or faxing
events, but the proportions of yellings or mutterings that are
described as communicative transfers of possession of information are
much smaller, we suspect, than the proportions of emailings or faxings
that are.

    Thus, for communication verbs our hypothesis can be summarized in
this way.  We can use both dative NP and dative PP syntax to express
communications, viewed as transfers of possession of information, but
the prototypical dative verb of communication, {\it tell}, is heavily
biased toward the dative NP construction.  Now communication is more
likely to be described in discourse about faxing, emailing, and other
events involving means of communication than in discourse about
whispering, yelling, mumbling, barking, and muttering.  Hence, we are
more likely to judge verbs in the {\it mutter} class as unacceptable
with dative NP syntax than verbs in the {\it fax} class simply because
there are far fewer instances of mutterings, mumblings, and yellings
that are likely to be described as instances of tellings.


\section{What drives the dative alternation?}

For the many cases where the choice of dative syntax is not determined
by meaning, what drives the dative alternation?  

     A simple Optimality Theoretic (OT) model of the dative
alternation can be based on two conflicting constraints on syntactic
structure, a faithfulness constraint requiring distinct marking of the
recipient role and an economy constraint penalizing syntactic
structure:\footnote{We
set aside additional constraints which would be needed to choose
between morphological (case) and syntactic (adpositional) dative
marking.}

\begin{quote}
    {\faith}: Express the recipient role of a verb
    with distinct marking (case or adposition)

    {\struct}: Avoid syntactic structure (here: *PP)\medskip
\end{quote}

{\small 
\begin{tabular}{|r|c|c|c|}\hline
                  & \struct & \faith & {\sc other}\\
                  &         &        & {\sc constraints}\\\hline
give them cheques &         &    *    &            \\\hline
give cheques to them &   *   &        &            \\\hline
give them-\dat\ cheques-\acc &  &       &    *        \\\hline
\end{tabular}
 }
\bigskip

\noindent This model assumes the same input for each candidate set;
the choice of syntax is always relative to a given meaning to be
expressed, and as we have seen the range of verbs and idioms allowing
overlapping meanings with alternative dative syntax is much greater
than has been appreciated.

    If only nonlinguistic factors were involved, the choice between
these alternants would be unpredictable from properties of grammar.
In this case stochastic (`noisy') evaluation of the structural
constraints, producing variable ranking, would be a sufficient model
(Boersma 1998, Boersma and Hayes 2001).  In OT with stochastic
evaluation the variable rankings of \struct\ and \faith\ produced by
noisy evaluation will lead to constraint reversals at a frequency
which is a function of the distance between the constraints on the
continuous ranking scale.  Given variable ranking normally distributed
around a mean, the closer together the constraints are, the more the
reversals, and the more variable the outputs.

\begin{figure}[htbp]\caption{Constraint ranking on a continuous scale with
    stochastic evaluation:}
\begin{center}
\includegraphics[width=4in]{dative-gaussian.eps}
\end{center}
\end{figure}

\sent{
\hfil\begin{tabular}{|lr|c|c|}\hline
            &      & \struct & \faith \\\hline
 \R &give them cheques &         &    *    \\\hline
  &give cheques to them &   *!   &        \\\hline
\end{tabular}

\hfil\begin{tabular}{|lr|c|c|}\hline
            &      & \faith & \struct \\\hline
 &give them cheques &    *!     &        \\\hline
\R  &give cheques to them &      &  *      \\\hline
\end{tabular}
 }

    An OT grammar with stochastic evaluation can generate both
categorical and variable outputs.  Categorical outputs arise when
crucially ranked constraints are distant.  As the distance between
constraints increases, interactions become vanishingly rare.  (A
distance of five standard deviations ensures an error rate of less
than 0.02\%.)\footnote{See Boersma and Hayes (2001: 50).  Units of
measurement are arbitrary.  With standard deviation = 2.0, a ranking
distance of 10 units between constraints is taken to be effectively
categorical.}



\subsection*{The \textsc{*NP Pro} Constraint}

But there are linguistic constraints on the dative alternation. 
One is the \nppro\ constraint widely cited in English linguistics.
Personal pronouns, but not demonstrative or indefinite
pronouns,are avoided when following lexical NPs if both are objects:

\sent{\begin{tabular}[t]{lll}
Erteschik-Shir (1979: 452): & Collins (1995: 39): & Kay  (1996):\\
John gave it to Mary.  & \bstar Tom gave an aunt them. &\bstar She
gave John it\\
\bstar John gave Mary it. &  Tom gave them to an aunt. &I gave John that.\\
                          &                           &She gave John one.\\
\end{tabular}
 }

\noindent A similar constraint appears in other languages, including
the Bantu language Lunda (Kawasha 2002).\footnote{Also Lillooet (van
Eijk 1997).}  In Lunda ditransitives with {\it give}, the recipient NP
object precedes the theme NP object:

\sents{
\item {\small \longgloss{3}
{N-\'e-enk-a & k\'ansi & {\bf mu-k\'anda}}
{i\,{\sc sg}-{\sc past}-give-{\sc fv} & 1.child & 3.book}\\
\nogloss{`I gave the child a book.'}
\item \longgloss{3}
{\bstar N-\'e-enk-a & {\bf mu-k\'anda} & k\'ansi }
{i\,{\sc sg}-{\sc past}-give-{\sc fv} & 3.book & 1.child }\\
\nogloss{`I gave the book a child.'}}
}

\noindent A pronominal secondary object can cliticize to the verb, but
only if the primary object is not a lexical NP:  %[check extraction]

\sents{
\item {\small \longgloss{2}
{\bstar N-e-enk-\'a={\bf wu} & {\red k\'ansi}}
{i\,{\sc sg}-{\sc past}-give-{\sc fv}={\sc pro}\,3 & 1.child}\\
\nogloss{`I gave it to the child.'}
\item \longgloss{1}
{N-e-{\red mw}-ink-\'a={\bf wu}}
{i\,{\sc sg}-{\sc past}-1\,{\sc obj}-give-{\sc fv}={\sc pro}\,3}\\
\nogloss{`I gave it to him.'}}
 }

\noindent Contrast a pronominal locative clitic in Lunda:

\sents{
\item {\small \longgloss{3}
{W-a--tentek-\'a & mali & {\bf mu-chisweki}.}
{1\,{\sc subj}-{\sc past}-put-{\sc fv} & 6.money & 8-7.cupboard } \\
\nogloss{`He put money in the cupboard.'}
\item \longgloss{3}
{\bstar W-a--tentek-\'a  & {\bf mu-chisweki} & mali.}
{1\,{\sc subj}-{\sc past}-put-{\sc fv} & 8-7.cupboard & 6.money  } \\
\nogloss{`He put in the cupboard money.'}
\item \longgloss{2}
{W-a--tentek-\'a={\bf mu} & {\red mali}}
{1\,{\sc subj}-{\sc past}-put-{\sc fv}={\sc pro}\,18 & 6.money} \\
\nogloss{`He put money in there.'}
\item \longgloss{1}
{W-a-{\red y\'{\i}}-tentek-\'a={\bf mu}}
{1\,{\sc subj}-{\sc past}-6\,{\sc obj}-put-{\sc fv}-{\sc pro}\,18}\\
\nogloss{`He put it in there.'}}\medskip
 }

 Ranking the \nppro\ constraint above all constraints which favor the
ditransitive NP NP dative will eliminate [V NP Pronoun] structures
from the output of the grammar:

\sent{
\hfil\begin{tabular}{|lr|c|c|c|}\hline
            && \nppro     & \struct & \faith \\\hline
 &give Mary them & *!  &       &    *    \\\hline
\R  &give them to Mary &  & *   &        \\\hline
\end{tabular}
 }
\medskip

\noindent For example, if \struct\ and \faith\ are ranked closely
enough together to create a threshhold of linguistically unpredictable
alternation through noisy evaluation, \nppro\ can be ranked so much
higher above both that the result is effectively a categorical absence
of variation for \mbox{\sc [Verb NP Pronoun]} inputs.

     In fact, however, the avoidance of NP pronoun sequences appears
to be gradient and not categorical in English.  It is true that the
parsed Switchboard corpus has no examples of ditransitives with the NP
Pronoun sequence.  But the following examples are representative of
those found in active use on the much larger corpus of web
documents.\footnote{The WWW is estimated to contain 47,000,000,000
words of English as of February 2000 (Baayen 2003, citing
Grefenstette 2000).}

\sent{\textsc{V NP Pronoun}\medskip


  Note: I don't give children peanut butter until they are 3 years old
  since it is recommended not to {\bf give children it} to avoid possible
  allergies.\\ {\small  Pratt's Children's Recipes \& Links
  www.fastq.com\slash $\sim$jbpratt\slash recipes\slash
  children\-recipes.\-html} \medskip

  You should never give out your address or phone number online and
  you should never {\bf send someone them} in the mail either.\\
  {\small www.girlpower.gov\slash girlarea\slash sciencetech\slash
  web\slash step1.htm} \medskip

  Per[c]eptions about God's absence are due to our lack of {\bf showing
  people him} through our life.\\ {\small
  christian-bookstore.net\slash xml031021436X\slash} \medskip

  Mega Blast beam: This is kakuri's strongest ki attack only he has
  what it takes to know how to use it he can {\bf teach people it} but
  it takes at least 2 years\\ {\small
  www.angelfire.com\slash ak4\slash DBZRPGgame/Kakuri\_s\_sheet.html} \medskip

  Please follow these simple rules and {\bf teach your children them},
  however most dogs are friendly.\\ {\small  Life4Paws-Knowledge Sharing,
  www.life4paws.org\slash sevenrules.htm} \medskip

  Second graders finished their underwater scenes and are very proud
  of these. They could not wait to {\bf show their parents them} and can't
  wait to bring them home.\\ {\small  cowlishaw.ipsd.org\slash specials.htm}

 }

\noindent Our model can very easily capture both the categoricity and the
gradience of the constraint across different languages, 
different varieties of English, or different speakers.

\begin{figure}[htbp]\caption{Categorical and gradient effects of \nppro:}
\begin{center}
\includegraphics[width=4in]{dative.eps}
\includegraphics[width=4in]{dative-grad.eps}
\end{center}
\end{figure}

     A more stringent version of this constraint is active for some speakers,
who avoid pronouns in second object position even when the primary
object is also pronominal:

\sent{
\nogloss{\bstar \ldots gave her it  (Erteschik-Shir 1979: 452)}\\
\nogloss{ \ldots gave him it  (Hawkins 1994: 312)}
}

\noindent Hawkins' (1994) grammaticality judgments reflect the less stringent
constraint, which only prohibits a secondary object pronoun in the
presence of a primary object lexical NP.  Only the less stringent
constraint is active in the Lunda data as well:

\sent{
Lunda: \\
\longgloss{1}
{N-e-{\bf mw}-ink-\'a={\bf wu}}
{i\,{\sc sg}-{\sc past}-1\,{\sc obj}-give-{\sc fv}={\sc pro}\,3}\\
\nogloss{`I gave it to him.'}
}

\noindent These differences are easily captured by means of constraint ranking:

\sent{

English (Erteschik-Shir):\hfill  \nppro\ \outranks\ \propro\ \outranks\
\struct\medskip

English (Hawkins): \hfill   \nppro\ \outranks\ \struct\  \outranks\
\propro\medskip 

Lunda: \hfill \nppro\ \outranks\ \struct\  \outranks\ \propro\

}

\subsection*{Generalizing the \textsc{*NP Pro} Constraint}

     These constraints undoubtedly reflect a broader generalization
about ditransitive objects, which has been characterized as
``Receiver/Entity Differentiation'' by Collins (1995: 47).  When there
are two NP objects, their properties are sharply differentiated and
polarized on scales of discourse accessibility, definiteness,
pronounhood, and word length.  The increased differentiation of
ditransitives is graphically depicted in Figure \ref{collins-graph},
based on Collins' (1995: 43) tabular data for givenness
(`accessibility').\footnote{For Collins (1995) `new' means either
``nonrecoverable because introduced for the first time into the
discourse'' or ``already present in the discourse, but newly
identified.''  `Accessible' means ``recoverable, but less directly so
than for a given entity, because the entity has to be inferred, is
generally known, or was first mentioned some time ago (or some
combination of these) and is therefore not as salient in the discourse
as a given entity.''  `Given' is defined as ``directly recoverable
because either previously mentioned or referred to directly in the
speech situation (or, in some cases, both).''}
\begin{figure}[htbp]\caption{Receiver/Entity Differentiation (Collins
    1995)}\label{collins-graph} 
\includegraphics[height=2.9in, width=2.9in]{Collins-graph2_np-np-nrml.ps}
\includegraphics[height=2.9in, width=2.9in]{Collins-graph2_np-pp-nrml.ps}
\end{figure}

       Receiver/Entity Differentiation is succinctly described by
Collins (1995: 47) in this way:

\begin{quotation}
In the indirect object construction the communicative differentiation
between receivers and entities is acute (receivers are almost 14 times
more likely than entities to be given, and entities are over 90 times
more likely than receivers to be new; entity NPs are over three times
longer than receiver NPs; receivers are over 11 times more likely than
entities to be expressed as pronouns; and receiver NPs are about 4.5
times more likely than entity NPs to be definite).  In the
prepositional construction, by contrast, the differences in
communicative status between receivers and entities are milder: along
no dimension is the order of the difference greater than 1.5.
\end{quotation}

     An OT model for this kind of pattern exists in Harmonic Alignment
(Prince and Smolensky 1993, Aissen 1999, 2002), and it has been
extended with stochastic evaluation (Boersma 1998, Boersma and Hayes
2001) to model gradient harmonic alignment patterns in English syntax
by Bresnan, Dingare, and Manning (2001) and Dingare (2001).  Here we
will simply encapsulate the multidimensional family of constraint
subhierarchies produced by this model in the Double-Object Primacy
Constraint:

\vbox{\sent{
  {\sc Double-Object Primacy (OO-Primacy)}:

%\begin{quote}
 When both are objects, the receiver/possessor (strictly) dominates
  the entity on hierarchies of informational prominence, and the
  entity (strictly) dominates the receiver/possessor on the reversed
  hierarchies:

\begin{center}

  Given \morharm\ Accessible \morharm\ New 

  Definite \morharm\ Indefinite

  Shorter \morharm\ Longer 

  Pronoun \morharm\ Noun  

   \ldots
\end{center}
%\end{quote}
}}


\paragraph{A side note.}

    The phenomenon of scope-freezing in ditransitives (Aoun and Li
1989, Bruening 2001) is probably an instance of the same
generalization.  There is indeed a strong preference for the first NP
in the double NP construction to scope over the second, compared with
the {\it to}-dative and the passive, and it may derive from this
increased sharpness of informational primacy of the primary object
over the secondary object:

\sents{
\item Ozzy gave a (different) telescope to each girl. \hfill  each
  $>$ a
\item Ozzy gave a (different) girl each telescope.  \hfill *each $>$ a
\item A (different) girl was given each telescope. \hfill each $>$ a
 }

\noindent Notice that this effect is also somewhat gradient, in that
if the second NP can be made accessible enough in context, scope reversal
seems grammatically possible:\footnote{Thanks to Ivan Sag for
  constructing these examples, and to Beth Levin and Peter Sells for
helpful  suggestions.} 

\sents{
\item My book collection had odd gaps in it. For example, I found that
    over the course of 20 years of teaching I had given one student
    or another each of my personal copies of \textit{Sense and
    Sensibility}. \hfill each $>$ a
\item It seemed likely that Bush had promised each campaign contributor
       a certain tax exemption(, which is ordinarily available only to
      foreign nationals). \hfill \mbox{a $>$ each}

}

   Bruening's (2001) syntactic analysis of the scope-freezing phenomenon
is independently motivated by by claimed grammatical impossibility of
{\it give} NP NP idioms to alternate, which we have seen is not not
empirically supported.  Thus, it seems desirable to explore a more
general informational approach to these semantic interactions with the
dative constructions.\footnote{Bruening (2001) also analyzes
superiority effects and antecedent-contained deletion which he
correlates with the scope-freezing phenemenon in ditransitives.
Though these topics lie outside the scope of the present study, we can
point to previous work suggesting an optimization approach to operator
binding, which straightforwardly extends to superiority effects
(Bresnan 1998). }


\subsection*{Core/Noncore Harmony}
 
     While improperly differentiated ditransitives are penalized
by OO-Primacy, there are also constraints penalizing disharmonic
{\it to}-dative constructions.

      Previous corpus studies of the dative alternation have found
evidence of a skewing of animate, definite, short, pronominal, and
given arguments toward the dative NP position and away from the dative
PP position (Giv\'on 1984; Thompson 1990, 1995; Collins 1995).
Giv\'on (1984) characterizes the dative NP as a grammaticalized
``secondary topic'' and Thompson (1990: 241, 1995: 158) interprets the
findings as evidence that the dative NP argument has more subject-like
informational properties (`topicworthiness') than the dative PP
argument.\footnote{Subject-like syntax for the NP-dative is embodied
in some syntactic derivations of datives as well (Larson 1988; cf.\
Jackendoff 1990, Larson 1990).}

     In corpus studies of the choice between active and passive, the
distribution of local persons has been found to be skewed toward the
subject argument and away from nonsubject arguments in English
(Estival and Myhill 1988, Bresnan, Dingare, and Manning 2001, Dingare
2001, ao).  If the dative NP shares informational properties with
subjects, as has previously been hypothesized, then we would expect to
find local persons harmonically aligned with the core dative NP
argument and nonlocal persons with the noncore dative PP argument.
This is the hypothesis we chose to investigate in the current study.

      In the Switchboard corpus, dative NPs of all types (pronominal
and nominal) are more frequent than dative PPs of all types, as we
already observed in Table \ref{allVerbs}: 83\% ($n=517$) of all the
dative arguments collected are expressed as dative NPs, and only 17\%
($n=104$) as dative PPs.  If we split up these dative arguments by
person, we find that the distribution of local (first and second)
persons is skewed toward dative NPs while the distribution of nonlocal
(third) persons is skewed toward dative PPs.  See Table \ref{big-person}.

\begin{table}[htbp]\caption{Person by Dative Argument Type in {\sc
      Switchboard}} \label{big-person}
\label{allVerbs-person}
\begin{center}
\begin{tabular}{lrr}
                    &  dative NP  &   dative PP\\\hline

  1$^{st}$, 2$^{nd}$ person &   296   &  20  \\
  3$^{rd}$ person          &    220      &  83 \\\hline
\end{tabular}\bigskip

  All alternating dative verbs, pronominal and nominal 3$^{rd}$
  persons 

Fisher's exact test, p(O$\ge$E) $<$ 0.000
\end{center}
\end{table}

       However, the property we are interested in, person, is highly
correlated with many other properties: for example, pronouns are
short, definite, and tend to be given, and local person pronouns are
in addition animate and seem nearly always given in conversations.  We
know that there is an effect of weight (correlated with word-length)
on the dative alternation (Thompson 1990, 1995; Hawkins 1994:
212--213, 311--313; Collins 1995; Wasow 1997, 2002).  Since local
person pronouns are all short while nonlocal person NPs are longer on
average, the weight or length effect would be in the same direction as
the person effect: the shorter would tend to precede the longer.  That
would bias local-person Recipients toward dative NP position adjacent
to the verb and away from dative PP position following the often
heavier or longer Themes.  Hawkins (1994) argues that apparent effects
of topicality or givenness can be better explained in terms of
syntactic weight and processing.\footnote{Arnold et al.\ (2000) and
Wasow (2002) have already shown that weight and informational status
have distinct effects on ordering.}  Would weight be sufficient to
explain the apparent harmonic alignment effects we observe in Table
\ref{allVerbs-person}?

       To address this issue, we controlled for length by excluding
examples which could be independently explained by the principle of
end weight: in ditransitives, the theme was restricted to one lexical
word (allowing for nonlexical elements such as determiners).  We also
restricted our search to personal and demonstrative pronouns in order
to control for pronominality.  To discount an effect of OO-Primacy in
this count, examples with personal pronoun themes were also excluded.
As before, all non-alternating verbs and expressions were eliminated;
this includes all concealed question uses of {\it tell} and 17
occurrences of {\it tell you what} (see n.\ \ref{tell-you-what}), as
well as ditransitives that do not alternate with {\it to} datives,
such as benefactives.   But in addition, 12 occurrences of the
sentential adverbial {\it to tell you the truth} were
eliminated.\footnote{Interestingly, this use of {\it to tell you the
truth} is not a fixed expression, but itself shows alternative dative
syntax.  It is not unnatural to say ``To tell the truth to you'', if
there is some reason to stress ``you''; at any rate, adverbial uses of
this expression with dative PP syntax are found on the internet;
several are poetic, perhaps exploiting the metrical possibilities of
this paraphrase:

\begin{quote}
  Let's see\ldots Bethany, {\bf to tell the truth
  to you}, I can't really remember. \\
  callistawolf.com\slash KymKyhgQ4\slash ep1pt2\slash page2.html\medskip

  You read your poetry I never understood. {\bf To tell the truth to you}, it
  was never any good.\\
  www.eightyproofsoul.com\slash LyricsPinkiePromises.html\medskip

  {\bf To tell the truth to you}, my love. when your comments are caging. the
  space that awaits me, \ldots\\
  www.racoon.com\slash herpes\slash biopage\slash MegAnn\%20poem.html
\end{quote}
  If the twelve excluded examples of {\it to tell you the truth} were
added to our data in Table \ref{other-verbs}, they would add support
to the harmonic alignment effect  (Fisher's exact test, p(O$\ge$E) $<$ 0.02).}

    Under these stringent conditions the verb {\it give} shows a total
absence of pronominal PP recipients for cases with non-pronominal
themes, and therefore any effect of person cannot be seen.  We
therefore eliminated {\it give} from the count.  For the remaining
verbs, first and second persons are still skewed toward the dative NP
position compared to the third person, as shown in Table
\ref{other-verbs}.
\begin{table}[htbp]\caption{Person by Dative Argument Type in {\sc
      Switchboard}} 
\label{other-verbs}
\begin{center}
\begin{tabular}{lrr}
                    &  dative NP  &   dative PP\\\hline

local pronoun             &    41       &      5  \\
nonlocal pronoun          &    19       &     8 \\\hline
\end{tabular}\bigskip

  Alternating dative verbs other than {\it give} with personal and
    demonstrative pronoun Recipients, less personal pronoun Themes and
    controlling for end weight.

Fisher's exact test, p(O$\ge$E) $<$ 0.05
\end{center}
\end{table}

   Our conclusion is that there is an effect of harmonic
alignment of person with the syntactic argument type (Core
$=$ NP or Noncore $=$ PP) of the recipient.  We now introduce the
constraint \harmlocal\ in our model:

\sent{{Core/Noncore Harmony for Person:}
  \label{harmony} 

 \textsc{Harmony(1,2)}:   *NP\sub{Noun} \& *PP\sub{1,2\,Person}

 }

    This harmony constraint can be formally derived by standard
methods in OT by harmonically aligning the pronominality and person
hierarchies with the structural hierarchy of Core and Noncore
arguments \exno{1} and making use of constraint conjunction.  These
formal techniques are now familiar in syntax from the work of Aissen
(1999, 2002) and others, and we will not repeat the construction here.
We note that animacy and other harmonies may also play a role in the
dative alternation in spoken English, but the parsed Switchboard
corpus is not annotated for this category. With limited resources,
we restricted our attention to the category of person for the present
study.\footnote{The Edinburgh-Stanford Link project Paraphrase
Analysis for Improved Generation aims to provide animacy, givenness,
and definiteness annotations to the parsed Switchboard corpus as one
of its goals (on-line, Edinburgh University: http:\slash\slash
www.hcrc.ed.ac.uk\slash stanford\slash).}

\sent{Prominence scales for Harmonic Alignment:

 Core \morharm\ Noncore  

   Pronouns \morharm\ Nouns

   Animate \morharm\ Inanimate

   Local (1$^{st}$, 2$^{nd}$) persons \morharm\ Nonlocal
  (3$^{rd}$) persons

}

      In summary, our (partial)  model of the constraints driving the dative
alternation now looks like this (again assuming ranking on a
continuous scale with stochastic evaluation):\footnote{The rankings
  shown are not meant to represent the mean rankings which derive the
input-output distributions under stochastic evaluation, but simply to
illustrate the dominance relations among constraints which will hold
under a portion of the effective rankings during the
evaluation.}

\sent{
\OO\ \outranks\ \harmlocal\ \outranks\ \faith\ \outranks\ \struct
 }

\noindent \OO, being near-categorical, is ranked above all the other
constraints.  \harmlocal\ favors the dative NP construction, while
\faith\ favors the dative PP construction, which is again opposed by
\struct.  The picture then is that what drives the dative alternation
are linguistic pressures to sharply differentiate double objects on
informational hierarchies, to prefer informationally prominent
elements in core argument positions, to faithfully mark the semantic
role of recipient, and to economize on syntactic structure.

     We now turn to the question or how lexical variation is accounted
for in this model.

\subsection*{Lexical Variation}

    The verb \textit{give} is often taken to be the prototypical
dative verb; it is the highest-frequency dative verb in the parsed
Switchboard corpus, constituting 42\% of all alternating dative verbs.
Yet it does not have the same distribution of syntactic argument types
as the pool of other dative verbs: {\it give} is used with dative NP
syntax in 87\% of its occurrences in the parsed Switchboard corpus,
compared to 81\% for the pool of other dative verbs, as we saw in
Table \ref{allVerbs}.

     The verb {\it give} also differs in having a higher percentage of
third persons among its pronominal dative arguments:

\vbox{\sent{All pronominal recipients\\(animate and inanimate, reflexive and
  nonreflexive) 

\begin{tabular}[t]{lrlrl}
  & \multicolumn{2}{l}{{\it give}}  &\multicolumn{2}{l}{other verbs}\\\hline
local person            &      153   &       & 148 & \\
nonlocal person         &      102   &(40\%) & 63  & (30\%)  \\\hline
totals:                 &      255   &       & 211 &   \\
\end{tabular}

}}

\noindent This difference is particularly striking with the pronoun
{\it it}: where 95\% (n$=$21) of dative NP occurrences of {\it it}
occur with {\it give}:

\sent{Distribution of recipient {\it it} for {\it give} and
other verbs

\begin{tabular}[t]{lll}
Other verbs & {\it give} & Total \\\hline
 1 (5\%)    & 21 (95\%)  &  22 (100\%)
\end{tabular}
}
\noindent  For inanimate recipients in general, {\it
give} diverges significantly from other dative verbs:

\sent{
Distribution of inanimate dative arguments (pronominal and nominal)
for {\it give} and other verbs

\begin{tabular}[t]{lrrr}
                & Other verbs &  {\it give}  &  Total \\\hline
    Dative NP:  &  2         &  25          &     27\\
    Dative PP:  &  9         &   1          &     10\\
\end{tabular}

Fisher's exact test, p(O$\ge$ E$\,|\,$O$\le$ E) $<$ 0.001
}

\noindent These differences undoubtedly reflect the greater
abstractness of the meaning of {\it give}, illustrated in examples like these: 

\sent{
 Um, but still, it gives {\bf it} some variety.

 but I'm going to give {\bf it} thumbs down.

 you know, give {\bf it} a great deal of thought,

 and you can add hamburger if you want to give {\bf it} a little more body
 }

\noindent All of these examples are paraphrasable with dative PPs
({\it give some variety to it, give thumbs down to that, give a great
deal of thought to it, give a little more body to it}), suggesting
that {\it give} here is not strictly polysemous, but generic or
abstract in meaning.\footnote{We assume that polysemy arises when
related meanings have distinct grammatical properties; such
grammatical differences are not present in these examples.}  This
reflects a parallel to the well-known generalization that high
frequency verbs tend to be more polysemous (Koehler 1986, cited in
Baayen and Tweedie 1998); here, the highest frequency verb has greater
semantic range.

     Interestingly, the extended senses derive from {\it give} together with
its Theme argument: to {\it give thumbs down} is to reject or disapprove;
to {\it give thought}; is to think about, to {\it give variety} is to
variegate, and so forth.   At the same time the recipient is preferred in the
dative NP position, separating the Theme from the verb.  These
facts with {\it give} run counter to the principles of semantic
connectedness (Hawkins 2000) and semantic distance (Bybee 1985)
which would favor closer syntactic groupings of semantically dependent
constituents and thus predict that the recipients would tend to appear in
dative PP position.  Evidently, other factors are overriding
the role of verbal or constructional semantics in harmonic person/type
alignment.  

    We can easily incorporate this lexical variation into our model by
distinguishing \faith\ for smaller lexical and constructional semantic
classes.\footnote{This idea was suggested to us by Stemberger's (2001)
work.}  Then the greater resistence to alternation of some dative
verbs will correspond to their more faithful observance of the
constraint to mark a Recipient role.  Consider the following
constraint ranking on the continuous scale:

\sent{\label{model1}  Incorporating lexical variation:

\hfil \OO\ \outranks
 
 \harmlocal\ \outranks\ \faithother\ \outranks\
\faithgive\ \outranks\ \struct
 }

\noindent In this ranking, \OO\ outranks the other constraints,
reflecting the near-categorical avoidance of double object NP Pronoun sequences.
The effect of \harmlocal\ on \faithgive\ is
greater than the effect of \harmlocal\ on \faithother\ (because of their
relative distance relations under stochastic evaluation).  This means
that {\it give} should obey \harmlocal\ to a greater extent than other
alternating verbs do.  That is, disharmonic cases of NP PP should be
rarer with {\it give} (for other verbs, disharmonic NP PPs are more
likely to result from reranking of \faithother\ with \harmlocal). This
implies that (i) the bias to express local recipients as objects
should be stronger for {\it give} than for other alternating verbs.

    At the same time, \faithother\ is affected less by \struct\
compared to \faithgive, so other verbs should be more rarely found in
NP NP with nonlocal recipients, again because of the relative distance
relations of the faithfulness constraints to \struct.  That is, the
percentage of NP NPs not motivated by \harm\ should be lower for other
alternating verbs than for {\it give} (with {\it give}, more NP NPs
will result from reranking of \faithgive\ and \struct, since the
distance is closer). This implies that (ii) nonlocal recipients
should be more often expressed as dative NPs with {\it give} than with
other verbs.

     Both of these consequences of this model are true.
(i) the bias to express local recipients as objects is stronger
for {\it give} than for other alternating verbs:

\sent{local person pronoun recipients:

\begin{tabular}[t]{rrr}
                                &  {\it give} &   other verbs\\\hline
                    NP          &    40       &  41  \\
                    PP          &     0       &   5  \\
\end{tabular}

Fisher's exact test, p(O$\ge$E) $<$ 0.04

 }

\noindent (ii) nonlocal recipients should is more often expressed as dative
NPs with {\it give} than with other verbs:

\sent{nonlocal person pronoun recipients:

\begin{tabular}[t]{rrr}
& {\it give} &   other verbs\\\hline
                    NP &             25    &     19 \\
                    PP &              0    &      8\\
\end{tabular}

Fisher's exact test, p(O$\ge$E) $<$ 0.03
 }

    Frequency differences among fine-grained semantic classes of verbs
can be incorporated into this model in the same way.  The ranking in
\exno{1} models the greater resistence of verbs of manner of speaking
({\it yell, mutter,} \ldots) and motional verbs of continuous
imparting of force ({\it carry, drag,} \ldots) to the dative NP
constructions.   These verbs are expected to alternate
less and therefore have more disharmonic dative PP constructions than
verbs of means of communication ({\it email, fax, \ldots}) and
motional verbs of instantaneous imparting of force ({\it throw, slap,
\ldots}).  In the following model, we use curley braces around an
exemplar verb as an abbreviation for a fine-grained class of dative
verbs which has a distinct \faith\ constraint; {\it give} is treated
as a singleton class for \faith, because of its prototypicality,
generic semantics, and overwhelming frequency.\footnote{Alternatively,
  these classes may be thought of as representing clusters of individuals in the
  constraint space.}
         
\sent{\label{model2} Incorporating more fine-grained classes:

\hfil \OO\ \outranks

{\sc Faith\sub{\{yell\}}, Faith\sub{\{drag\}}}
\outranks\ \harmlocal\ 
\outranks\ {\sc Faith\sub{\{fax\}}, Faith\sub{\{throw\}}}

\hfil  \outranks\ \ldots\ \outranks\ {\sc Faith\sub{\{give\}}(Rec)}
\outranks\ \struct 

}

For verbs that fail to alternate, such as the latinate class
\textit{donate, contribute,\ldots}, \faith\ can be ranked very high,
near-categorically far removed from the constraints which result in
dative NP syntax:

\sent{  Incorporating nonalternating classes:

\hfil \OO, \faithlatin\ \outranks

{\sc Faith\sub{\{yell\}}, Faith\sub{\{drag\}}}
\outranks\ \harmlocal\ 
\outranks\ {\sc Faith\sub{\{fax\}}, Faith\sub{\{throw\}}}

\hfil  \outranks\ \ldots\ \outranks\ {\sc Faith\sub{\{give\}}(Rec)}
\outranks\ \struct 

}

%[jb: comment on future work: derive rankings from frequency.  Add
%  preliminary GLA simulations here?]]

In sum, there is a gradation in frequency of alternation of
verb classes, from those that are categorically used with dative PPs
(most of the latinate class), to those that are rare with dative NPs
(motional verbs of continuous imparting of force, manner of speaking
verbs), through the more frequently ditransitive classes (other
motional and communication verbs), to the most frequently ditransitive
({\it give}). \textit{According to the model, the more frequently
ditransitive alternating verbs are more driven by the harmonic effects of
informational hierarchies (such as discourse accessibility, animacy,
pronominality and person), as well as economy.}

\section{Are gradient patterns linguistically significant?}

    Within our stochastic model of grammar, the structure of language
has remarkable plasticity.  The boundaries between categoricity and
gradience are fluid.  We therefore expect to find languages in which
the gradient but broadly motivated patterns we have discovered in
English are hardened into categorical rule systems.

    Several languages show categorical pronominal and person splits in
the dative alternation (Nikitina 2003), but we consider just one here.
Kanuri, a Nilo-Saharan language spoken in Nigeria, Niger, and
Cameroon, shows a person split across alternative dative structures
(Hutchison 1981).
In Kanuri with the verb {\it give} a non-local person recipient can be
expressed only in a postpositional phrase:

\sent{
\longgloss{2}
{\textit{shi-ro}&\textit{yik\textschwa na}}
{him-to& give-{\sc prf}}\\
\nogloss{`I gave (it) to him'}
 }

\noindent Note that it appears to be highly dispreferred to drop the third
person recipient:\footnote{  }

\sent{
\longgloss{1}
{\bqm \textit{yik\textschwa na}}
{0-give-{\sc prf}}\\
\nogloss{`I give (it) to him'}
 }

\noindent However, if the recipient is second or first person, it is normally
expressed as a direct object prefix on the verb:

\sent{
\longgloss{1}
{\textit{nj-ikin}}
{{\sc 2Sg.Obj}-give}\\
\nogloss{`I give (it) to you'}
 }

          This split is straightforwardly captured by the constraint
ranking shown in \exno{1}.

\sent{

     \harmlocal\ \outranks\  \faithgive \outranks\ \struct\medskip

\hfil\begin{tabular}{|lr|c|c|c|}\hline
& /give it to him/  & \harmlocal     & \faith & \struct \\\hline
 &him-to (it) give &    &      &  *     \\\hline
\R  &(it) him-give &    &   *!  &        \\\hline
\end{tabular}

\hfil\begin{tabular}{|lr|c|c|c|}\hline
& /give it to you/  & \harmlocal     & \faith & \struct \\\hline
 &you-to (it) give &  *!  &       & *       \\\hline
\R  &(it) you-give &      &  * &        \\\hline
\end{tabular}
 }\medskip

\noindent In Kanuri \harmlocal\ is ranked above \faithgive\, and so if the input
is `I give it to you', the candidate with the recipient expressed in a
postpositional phrase is excluded by highly ranked \harmlocal.  If the
input is `I give it to him', \harmlocal\ remains inactive (the
recipient is non-local person), and it is the variant with
postpositional expression of the Recipient (satisfying \faithgive\ that
wins in that competition.

     Finally, note that in Kanuri only the single verb meaning {\it
give} shows a dative alternation.  All other verbs express all
recipients, whether local or nonlocal persons by means of
postpostional phrases.  Thus a fuller model of the Kanuri constraint
system is shown in \exno{1}:


\sent{

   \faithother\ \outranks\  \harmlocal\ \outranks\  \faithgive
   \outranks\ \struct 

 }

     The Kanuri constraint system for the dative alternation resembles that
of English (\ref{model2}), but the Kanuri constraints will be spread
sufficiently far apart on the continuous scale to produce a
(near-)categorical person split.\footnote{Compare Bresnan, Dingare, and
Manning 2001 for a similar demonstration of gradient and categorical
person splits affecting active-passive choice.}   

     In English, the generalization that the most frequently alternating
dative verb is most driven by harmony is gradient; it Kanuri, it is
near-categorical.\footnote{Claims of categoricality must be carefully
evaluated for every language, taking into account both textual and
elicitation data, as is currently the best practice in field linguistics.
If the Kanuri generalization is itself quantitative, the frequency
differentials still support our model.  See Bresnan, Dingare, and
Manning (2001).}




\section*{Conclusion}

    In conclusion, we have shown that central evidential paradigms
that have been used to support both lexical semantic (Pinker 1989,
Krifka 2001) and constructional (Goldberg 1995, Kay 1996)
explanations for syntactic contrasts are not well founded empirically.
Some widely repeated reports of intuitive contrasts in grammaticality
appear to rest instead on judgments of pragmatic probabilities.
We have also shown that at least one type of informational
structure (specifically the hierarchy of person or speech act
participants) exerts an effect on the dative alternation independently
of effects of length or weight, semantic role, and pronominality.  The
findings show that syntactic processing theories based on factors
correlated with length (e.g.\ Hawkins 1994) are not sufficient
explanations, though they may well be necessary to a full
understanding of the dative alternation.  Further, we have proposed a
unifying model of the person alignment phenomenon within the framework
of stochastic Optimality Theory, and have shown how lexical variation
can be incorporated into the model.  Within the framework of this
model, the constraint ranking for English implies that the most
frequently ditransitive alternating verbs should be the most driven by
informational harmony.  
%[jb: include? We have argued that the relative frequencies
%of alternation of different verbs could depend on  their
%frequencies of occurrence in the input to a statistically sensitive
%learning algorithm (the GLA).]
Finally, we have shown that the same
pattern of person/argument alignment that appears gradiently in the
English dative alternation appears near-categorically in Kanuri, as
our model leads us to expect.

  We note in conclusion that the stochastic OT model used in the present
study belongs to a family of new probabilistic approaches that permit unifying
explanations of categorial and gradient effects in
syntax.\footnote{Other promising models include partially ordered
constraint ranking (Anttila and Fong 2003 and references) and the
wide range of probabilistic models discussed in Bod, Hay, and Jannedy
(eds) (2003) and Manning (2003).}

\section*{References}


\leftskip =\parindent
\parindent =-\parindent
\parskip =\medskipamount

\indent

Aissen, Judith. 1999.  Markedness and Subject Choice in Optimality
Theory. {\it NLLT} 17: 673--711.

Aissen, Judith. 2002. Differential object marking:
Iconicity vs.\ economy. {\it NLLT}.  To appear.

Anttila, Arto and Vivienne Fong.  Variation, ambiguity, and noun
classes in English.  On-line, Rutgers Optimality Archive:
http:\slash\slash roa.rutgers.edu. 

Aoun, Joseph and Yen-hui Audrey Li. 1989. Constituency and scope.
\textit{Linguistic Inquiry} 30: 275--343.

Arad, M. 1998.  VP-Structure and the Syntax-Lexicon Interface.
London:  University College London Ph.D.\ dissertation.

Arnold, Jennifer, Thomas Wasow, Anthony Losongco, \& Ryan
Ginstrom. 2000.  Heaviness vs.\ newness: The effects of complexity and
information structure on constituent ordering. {\it Language} 76: 28--55.

%Ariel, Mira.  1990.  {\it Accessing NP Antecedents.} London: Routledge,
%Croom Helm Linguistics Series.

Baayen, R.\ H.\ 2003.  Probabilistic approaches to morphology.  In
\textit{Probabilistic Linguistics}, ed.\ by Rens Bod, Jennifer Hay and
Stefanie Jannedy, 229--287.  Cambridge, Massachusetts: The MIT PRess.

Baayen, R.\ H.\ and Tweedie, F.\ J.\ 1998.  Mixture models and word
frequency distributions.  Abstracts of the ALLC/ACH Conference,
Debrecen, July 1998.

Bod, Rens, Jennifer Hay and Stefanie Jannedy.  (2003).
\textit{Probabilistic Linguistics}.  Cambridge, Massachusetts: The MIT PRess.

Boersma, Paul. 1998.  \textit{Functional phonology: Formalizing the
Interactions between Articulatory and Perceptual Drives}.  Amsterdam:
University of Amsterdam Ph.D.\ dissertation.  On-line, University of
Amsterdam:  http:\slash\slash www\-.fon\-.hum\-.uva\-.nl\slash paul\slash.

Boersma, Paul and Bruce Hayes. 2001. Empirical tests of the Gradual
Learning Algorithm.  \textit{Linguistic Inquiry} 32: 45--86. 

Boersma, Paul and David Weenink. 2000.  Praat computer program.  On-line,
Institute of Phonetic Sciences, University of Amsterdam:
http:\slash\slash www.\-fon.\-hum.\-uva.\-nl\slash praat\slash .

Bresnan, Joan. 1998.  Morphology competes with syntax: Explaining
    typological variation in weak crossover effects.  In {\it Is the
    Best Good Enough?  Optimality and Competition in Syntax}, edited
    by Pilar Barbosa, Danny Fox, Paul Hagstrom, Martha McGinnis, and
    David Pesetsky, 59--92. Cambridge, Massachusetts: The MIT Press
    and MIT Working Papers in Linguistics.

Bresnan, Joan, Shipra Dingare, and Christopher D. Manning. 2001.
Soft constraints mirror hard constraints: Voice and person in English
and Lummi.  {\it Proceedings of the LFG '01 Conference}. CSLI
Publications, on-line. 

Bruening, Benjamin.  QR obeys Superiority:  Frozen scope and ACD.
\textit{Linguistic Inquiry} 32: 233--273.

Bybee, Joan L.  1985. \textit{Morphology: A Study of the Relation
  between Meaning and Form}.  Amsterdam: John Benjamins.

Collins, Peter. 1995.  The indirect object construction in English: an
informational approach.  {\it Linguistics} 33: 35--49.

Davidse, Kristina.  1996a.  Functional dimensions of the dative in
English.  \textit{The Dative, Volume 1: Descriptive Studies}, edited by 
William Van Belle and Willy Van Langendonck, 289--338.  Amsterdam:
John Benjamins.

Davidse, Kristina. 1996b.  Ditransitivity and possession.  In
\textit{Functional Description: Theory in Practice}, ed.\ by Ruqaiya
Hasan, Carmel Aoran, David Butt, 85--144.  Amsterdam: John Benjamins.

%Davidse, Kristina.  1998a.  The dative as participant role versus the
%indirect object.  In \textit{The Dative: Volume 2.\  Theoretical and
%  Contrastive Studies}, ed.\ by Willy Van Landendonck and William Van Belle,
%143--184. 

Davidse, Kristina.  1998.  Agnates, verb classes, and the meaning of
construals.  The case of ditransitivity in English.  \textit{Leuvense
  Bijdragen.  Leuven Contributions in Linguistics and Philology}  87:
281--313. 

Davis, Anthony. 1997.  Lexical Semantics and Linking
in the Hierarchical Lexicon.  Stanford: Stanford University Department
of Linguistics Ph.D.\ dissertation.

Dingare, Shipra. 2001.  The effect of feature hierarchies on
frequencies of passivization in English.  Stanford University
Department of Linguistics M.A. thesis.  On-line, Rutgers Optimality
Archive: http:\slash\slash roa.rutgers.edu.

Erteschik-Shir, Nomi.  1979.  Discourse constraints on dative movement.
In \textit{Discourse and Syntax.  Syntax and Semantics}, vol.\ 12,
ed.\ by Talmy Giv\'on, 441--467.  New York: Academic.

Estival, Dominique and John Myhill. 1988. Formal and functional
aspects of the development from passive to ergative systems.  In
\textit{Passive and Voice}, ed.\ by Masayoshi Shibatani, 441--491.  Amsterdam:
Benjamins. 

Giv\'on, Talmy.  1984.  Direct object and dative shifting: Semantic and
pragmatic case. In \textit{Objects.  Towards a Theory of Grammatical
  Relations,} ed.\ by Frans Plank, 151--182.  New York: Academic Press.

Godfrey J., E.\ Holliman and J.\ McDaniel. 1992. \textsc{switchboard}:
Telephone speech corpus for research and development. \textit{Proceedings of
  ICASSP-92}, San Francisco, pp.\ 517--520.

Goldberg, Adele. 1995.  \textit{Constructions: A Construction Grammar
  Approach to Argument Strucutre}.  Chicago: University of Chicago Press.

Green, Georgia. 1974.  {\it Semantics and Syntactic Regularity.}
Bloomington: Indiana University Press.

Grefenstette,G.  2000.  Very large lexicons from the WWW.  Paper
presented at the University of Brighton, U.K.

Gropen, Jess, Steven Pinker, Michelle Hollander, Richard Goldberg, and
Ronald Wilson.  1989.  The learnability and acquisition of the dative
alternation.   \textit{Language} 65: 203--257.

Halliday, M.\ A.\ K. 1970. Language structure and language function.
In \textit{New Horizons in Linguistics}, ed.\ by John Lyons,
140--165.  Harmondsworth, Middlesex: Penguin.

Harley, Heidi.  1996.  If you {\it have}, you can {\it give}.  {\it WCCFL} 15.

Harley, Heidi. In press.  Possession and the double object
construction.  In \textit{Yearbook of Linguistic Variation II}, ed.\
by P.\ Pica and J.\ Rooryck.  Amsterdam: John Benjamins.

%Haspelmath, Martin. 2002. Explaining the Ditransitive Person-Role
%Constraint: A usage-based approach. (revised version, August 2002.
%Available on his website.

Hawkins, Jack. 1994.  {\it A Performance Theory of Order and Constituency.}
Cambridge University Press.

Hawkins, Jack. 2000. The relative order of prepositional phrases in
English:  Going beyond Manner--Place--Time.  \textit{Language,
  Variation, and Change} 11: 231--266.

Herslund. Michael. 1986. The double object construction in Danish.  In
{\it Topics in Scandinavian Syntax}, ed. by L.\ Hellan and K.\ Koch
Kristinsen, 125--147.  Dordrecht: D.\ Reidel.

Hutchison, John P. 1981  \textit{A Reference Grammar of the Kanuri
  Language}. Madison: African Studies Program. 

Jackendoff, Ray. 1990.  On Larson's treatment of the double object
construction.  {\it Linguistic Inquiry} 21: 427--456.

Kawasha, Boniface.  2002.  Grammatical relations and relativization in
Lunda.  {\it JALL} 23: 31--62.

Kay, Paul. 1996.  On-line lecture on construction grammar,
http:\slash\slash www.icsi.berkeley\-.edu\slash $\sim$kay\slash
bcg\slash 5\slash lec05\-.html.

Koehler, R. 1986. {\it Zur linguistischen Synergetik: Struktur und Dynamik
der Lexik}.  Bochum: Brockmeyer.

Krifka, Manfred. 2001.  Lexical Representations and the nature of the
Dative Alternation. University of Amsterdam, November 9, 2001. Handout
available on his webpage.

Larson, Richard K.  1988. On the double object construction.
\textit{Linguistic Inquiry} 19: 335--391.

Larson, Richard K.  1990. Double objects revisited: reply to
Jackendoff.  \textit{Linguistic Inquiry} 21: 589--632.

Levin, Beth. 1993.  \textit{English Verb Classes and Alternations.  A
  Preliminary Investigation}.  Cambridge, Massachusetts: The MIT Press.

Levin, Beth and Malka Rappaport Hovav. 2002. What alternates in the dative
alternation?.  The 2002 Conference on Role and Reference Grammar: New
Topics in Functional Linguistics: The Cognitive and Discoursive
Dimension of Morphology, Syntax and Semantics, Universidad de La
Rioja, Logrono, Spain, July 27--28, 2002.  Available on Levin's webpage.

%Maling, Joan.  2001.  Dative: the heterogeneity of the mapping among
%morphological case, grammatical functions, and thematic roles.
%\textit{Lingua} 111: 419--464.

Manning, Christopher D. 2003.  Probabilistic syntax. In
\textit{Probabilistic Linguistics}, ed.\ by Rens Bod, Jennifer Hay and
Stefanie Jannedy, 289--341.  Cambridge, Massachusetts: The MIT PRess.

Marantz, Alec.  1993.  Implications of asymmetries in double object
constructions.  In \textit{Theoretical Aspects of Bantu Grammar}, ed.\
by Sam A.\ Mchombo, 
113--150.  Stanford: CSLI Publications.

Marcus, Mitchell, Beatrice Santorini, and May Ann
Marcinkiewicz. 1993. Building a large annotated corpus of English: The Penn
treebank. \textit{Computational Linguistics} 19: 313--330.

%Newmeyer, Frederick.  2003.  Grammar is grammar and usage is usage.
%Presidential address.  LSA, January 2003, Atlanta, Georgia.

Nikitina, Tatiana.  2003.  Harmony constraints on clitic sequences.
Stanford: Stanford University Department of Linguistics {\sc ms}.

Oehrle, Richard T. 1976.  The grammar of the English dative
alternation.  Cambridge, Massachusetts: MIT Department of Linguistics Ph.D.\
dissertation. 

Pesetsky, David. 1995.  \textit{Zero Syntax}. Cambridge, Massachusetts: The
MIT Press.

Pinker, Steven. 1989.  \textit{Learnability and Cognition.  The Acquisition of
Argument Structure.}  Cambridge, Massachusetts: The MIT Press.

Prince, Alan and Paul Smolensky. 1993. \ital{Optimality Theory:
Constraint Interaction in Generative Grammar}.  RuCCS Technical Report
\#2. Piscataway, NJ: Rutgers University Center for Cognitive Science.

Ransom, Evelyn N. 1977.  Definiteness, animacy, and NP ordering.  {\it
  BLS} 3: 418--429.

Ransom, Evelyn N. 1979.  Definiteness and animacy constraints on passive
and double-object constructions in English.  \textit{Glossa} 13: 215--240.

Smyth, R.\ H., D.\ Prideaux, and J.\ T.\ Hogan. 1979.  The effect of
context on dative position.  \textit{Lingua} 47: 27--42.

Snyder, Kieran.  2001.  What can the pragmatics of the double object
construction  tell us about their syntax?  {\it CLS} 37.

Speas, Margaret. 1990. \textit{Phrase Structure in Natural Language}.
Dordrecht: Kluwer.

Stemberger, Joseph. 2001. Overtensing within Optimality Theory.
On-line, Rutgers Optimality Archive: http:\slash\slash
roa.rutgers.edu.

Thompson, Sandra. 1990. Information flow and dative shift in English
discourse.  In {\it Development and Diversity, Language Variation Across
Space and Time}, ed.\ by Jerold A. Edmondson, Feagin Crawford, and Peter
Mu\"hlh\"ausler, 239--253..  Summer Institute of Linguistics, Dallas, Texas.

Thompson, Sandra. 1995.  The iconicity of ``dative shift'' in English:
Considerations from information flow in discourse.  In {\it Syntactic
Iconicity and Linguistic Freezes}, ed.\ by Marge E.\ Landsberg, Mouton de
Gruyter, Berlin, pp. 155--175.

Van Eijk, Jan.  1997. \textit{The Lillooet Language:  Phonology,
  Morphology, Syntax}.  Vancouver: University of British Columbia
  Press.        

Wasow, Tom.  1997. Remarks on grammatical weight, {\it Language Variation and
Change} 9: 81-105, 

Wasow, Thomas. 2002.  \textit{Postverbal Behavior.}  Stanford: CSLI
Publications. 

Zwicky, Arnold, 1971, In a manner
of speaking. \textit{Linguistic Inquiry} 2: 223--233.





\end{document}