Ditransitive construction vs. prepositional dative

Joshua Viau, in a paper Give = CAUSE + HAVE / GO: Evidence for Early Semantic Decomposition of Dative Verbs in English Child Corpora (pdf discusses a difference in verbs the take both the di-transitive construction (I gave her it) and the prepositional indirect object construction ("I give it to her").

Viau takes the two sentences

1	Kathy sent a steel drum to Chicago
2	*Kathy sent Chicago a steel Drum

Ignoring here the metonymic reading of "Chicago" as something akin to "the Chicago Office", Viaou note that the first, with the prepositional construction, is acceptable and the second, with a ditransitive form, is not. This difference is analyzed as the difference between the prepositional form encoding "Kathy CAUSE (a steel drum GOTO Chicago)" while the ditransitive form encodes "Kathy CAUSE (Chicago HAVE a steel drum). Since, according to the analysis, Chicago is not the sort of thing that can HAVE a steel drum, the ditransitive form is disallowed, although it is allowed as a location to which something can be sent.

Viau's primitive HAVE seems to be a stative verb; that is, it encodes something -AT- a -LOC- (using the dashes to distinguish Viau's primitives from those along the line I have discussed before). Viau does seem to argue this, encoding the English verb [BE [x HAVE y]], BE apparently being a primitive which marks states rather than events.

Viau also distinguishes the primitives GO and BECOME, Go apparently encoding movement (perhaps metaphorically), and BECOME appearing to encode some more complex change in state or condition; in the paper BECOME is used in the explication of get as [BECOME [x HAVE y]]. The paper also distinguishes, without much discussion, the prepositional dative as [X CAUSE [y GO z]] from causative verbs [x CAUSE [y BECOME XP_state]].

In the somewhat simpler, or lower level, analysis I set forth earlier, Viau's HAVE is an -AT- verb, encoding that the Theme (what is had) is -AT- (in some sufficiently abstract form) the LOC (the person how has).

But sent in sentences 1 and 2 above encodes a change in -LOC-. The sentences have an implicit prior condition as [NOT [steel drum -AT_ Chicago]] and the final, stated condition of [steel drum -AT- Chicago]. Thus it is an event, and thus must encode a -GOTO-. So the sentences both would, at their base, be [Kathy -CAUSE- [steel drum -GOTO- Chicago].

The primitive CAUSE seems to only accept events, not states. In my analysis *[Agent -CAUSE- [Theme -AT- Loc]] appears not to occur; there is the event -STAY-, which can be CAUSEd, as in [gravity -CAUSE- [book -STAY- table], but the plain -AT- appears not to occur.

Viau's construction [Kathy CAUSE [Chicago HAVE steel drum]] thus does imply a change in state - with BECOME in the analysis of the paper. it seems a more complete explication may be [Kathy CAUSE [BECOME [Chicago HAVE steel drum]]. Since, as it appears that CAUSE can only take events, the event of BECOMEing is implicit, but seems to be present.

My analysis is at a lower level that that of the paper, and Viau's HAVE would, in my analysis, a state, a Loc, of having, which would have the semantic and pragmatic restrictions on what sort of things can have other sorts of things. Thus, in the most abstract form, sent would be analysed as [x -CAUSE- [y -GOTO- y]]. The ditransitive form, which Viau argues requires that the resultant condition include [z HAVE y], would encode more explictly the resultant form of [y -AT- z], with the semantic and pragmatic considerations that lead Viau to conclude that Chicago cannot be interpreted as a possessor of a book (or steel drum) being separate form the underlying highly abstract structure of the verb.

Theme incorporating Verb in Spanish, but not English

In El Mundo, there was a story on 14 Jan 2011 that began La Fundación BBVA ha premiado al economista británico Nicholas Stern por un informe de 2006 en el que afirma que no combatir el cambio climático es más caro que reducir la emisión de gases de efecto invernadero. We are interested here in the main verb, ha premiado, from premiar. The verb means, "to present an award", so the sentence means The BBVA Foundation has presented an award to British economist Nicholas Stern for a 2006 report which states that not combating climate change would be more expensive than reducing emissions of greenhouse gases.

The same construction is in this Antena3 story of 22 Sept 2010: Cibeles ha premiado a Helbig por la colección que presentó el sábado que destacó por los abrigos de verano, los colores claros y el cuero. (Cibeles has presented an award to Heiberg for the collection presented Saturday that stood out for summer coats, light colors and leather.)

What we see here is that the basic structure of the report is [The BBVA Foundation] CAUSE [an award] GOTO [British ecomomist Nicholas Stern]...; but the Spanish verb premiar has incorporated the theme ("an award"), so that it does not appear explictly in the sentence, while in English we do not seem to have a verb "to award" which so incorporates the theme; we cannot say *The BBVA Foundation awarded British economist Nicholas Stern ....

In traditional English grammar, the presenter (the BVVA Foundation) is the subject, what is presented (an award) is the direct object, and the recipient (Nicholas Stern) is the indrect object; in the notation here the presentor it the AGENT, that which is presented is the THEME, and the recipient is the LOC.

In Spanish, the THEME is incorporated into the verb, bringing the LOC into the direct object position. This results in the case here, where a single verb (ha premiado) must be translated into a verb-noun phrase (has presented an award) in English, because the Spanish verb incorporates the noun, while the English verb sdoes not.

Two levels of Abstraction

There seems to be two different levels of abstraction in the use of a word. Take, for instance, words related to marriage. There are at the first level the use of these words to apply to people. In my taxonomy in previous posts, X is married can be understood as X AT married, or, in more idiomatic phrasing, X is in the state of being married. (For the purposes of this post I'll simply ignore the issues revolving around the tense and aspect.) X got married is something like X GOTO married, or X became married. X got divorced, means something similar to X LEAVE married, or X leaves the state of being married. Similarly spouse has the meaning AT married, with wife including the meaning AT woman and husband the meaning AT man in addition the AT married.

Wed, like got married, includes the meaning GOTO married, a wedding is a ceremony in which people GOTO married. With the meaning The minister married the couple, the meaning seems to include something along the lines of minister CAUSE (Couple GOTO married). Notice all of these seem to encode a kind of adjectival meaning (which in my taxonomy shows fits in as a kind of stative verb) of married.

But what of the word marriage. The various words have a meaning such as married above. In my taxonomy, it can be viewed as a stative verb with the LOC "married" and a THEME, in normal usage, of a person.

The word marriage, however, seems to be "up" a level of abstraction; it does not place people into a relation with "being married", either in, entering, leaving, or causing, but rather "looks down" on the meaning if "being married" and uses that meaning as a noun.

Thus in a sentence (from Wikipedia) such as "Marriage is a social union or legal contract between people that creates kinship.", the word "marriage" is the THEME, not a verb, and does not relate a person to the state of "being married", but rather talks about what it means to be married.

So it appears that there are (at least) two different levels of abstraction to a verb in the taxonomy, one where we use the word, and the other were we turn it into a noun and use the noun to refer to the meaning of the verb. There seems to be two different levels of abstraction in the use of a word. Take, for instance, words related to marriage. There are at the first level the use of these words to apply to people. In my taxonomy in previous posts, X is married can be understood as X AT married, or, in more idiomatic phrasing, X is in the state of being married. (For the purposes of this post I'll simply ignore the issues revolving around the tense and aspect.)

X got married is something like X GOTO married, or X became married. X got divorced, means something similar to A LEAVE married. Similarly spouse has the meaning AT married, with wife including the meaning AT woman and husband the meaning AT man in addition the AT married.

Wed, like got married, includes the meaning GOTO married, a wedding is a ceremony in which people GOTO married. With the meaning The minister married the couple, the meaning seems to include something along the lines of minister CAUSE (Couple GOTO married.

Notice all of these seem to encode a kind of adjectival meaning (which in my taxonomy shows fits in as a kind of stative verb) of married.

But what of the word marriage. The various words have a meaning such as married above. In my taxonomy, it can be viewed as a stative verb with the LOC "married" and a THEME, in normal usage, of a person.

The word marriage, however, seems to be "up" a level of abstraction; it does not place people into a relation with "being married", either in, entering, leaving, or causing, but rather "looks down" on the meaning if "being married" and uses that meaning as a noun.

Thus in a sentence (from Wikipedia) such as "Marriage is a social union or legal contract between people that creates kinship.", the word "marriage" is the THEME, not a verb, and does not relate a person to the state of "being married", but rather talks about what it means to be married.

So it appears that there are (at least) two different levels of abstraction to a verb in the taxonomy, one where we use the word, and the other were we turn it into a noun and use the noun to refer to the meaning of the verb.

I, me, and overcorrection

The Vancouver Olympics are over now. The Anthem of the Vancouver Olympics (lyrics here) has the repeated lines:

I believe together we’ll fly
I believe in the power of you and I

Note that last line - I believe in the power of you and I. Shouldn't it be "... you and me."? Well, yes. Among those who care about such things, this has been critized as wrong (You can Google "I believe in the power of you and I" - with the quotes).

Some say that the rhyme with "fly" necessitates the error, that the artistic trumps correctness. But that doesn't really answer the question. The lyricist obviously thought this was, if not correct, close enough to correct that he could get away with using it.

Grammactically, this is called "overcorrection", when a correct grammatical form is changed to an incorrect form because it is related closely to a form which is often used incorrectly, and a change to the incorrect form to the correct form is overapplied to cases where the "correction" is incorrect (is that sentence complicated ehough?). In English "I" and "me", by the usual version of "proper grammar", are often misused. "Me and Joe went to the store", for instance, instead of the "correct" "Joe and I went to the store". This sets up in the mind a tendency to feel that "me" is used too often and "I" not enough, and this tendency displays itself in changing "me" to "I" inappropriately - "Mary gave the food to Joe and me" becomes ""Mary gave the food to Joe and I", for instance, and, as in the song, "the power of you and me" becomes "the power of you and I".

What seems to be happening is that in English, the primary form is the objective, and it gets changed by the brain, more or less at the last minute, to the nominative form. The brain forms the original sentence along the lines of "me go to the store" and, after the sentence has been formed but before it has been uttered, changes the "me" to "I".

Several things lead to this conclusion. First, something along this line has already happened in English. In the King James Bible of 1611, for example, the nominative case of "you" was "ye" ("ye" and "you" were only plural, the singular forms were then "thou" and "thee"). Note that it is the nominative form that disappeared, being taken over by the objective form. This suggests that the objective form is somewhat "stonger" or "more basic" in English than the nominative form.

Another thing that suggests this is what we might call "Tarzan English". Note it is "Me Tarzan", and not "I Tarzan", and "Me want .. " not "I want ...". However, we would not see in such "Tarzan" English "Give I ..."; it would still be "Give me ...". Similarly, it would be "Him Cheetah" rather than "He Cheetah", and "Give him ..." rather than "give he ...", and "Her Jane" rather than "She Jane".

This suggest that the objective forms ("me", "him", "her", "us", "them") are somehow "more basic" or "more fundamental" than the nominative forms, and these more basic forms are somehow corrected to the nominative forms later in the speech generating process. ("you" and "it" are the same in the nominative and objective cases, so the final result it the same: "You Jane")

What seems to, in normal informal English, the transformation of "me" to "I" is the presences of the pronoun alone in the normal pre-verb position. We do not normally see the use of "me" as the subject pronoun when it is alone: "Me will go to the store" is almost never seen in native English speakers, it is always "I will go to the store." However, the proceedure that changes the "me" to "I" seems to be sometimes "fooled" with nore complex subjects - will will see "Me and Joe will go to the store", or "Joe and me will go to the store". In this case is appears that the subject, being "Me and Joe", or "Joe and me" somehow doesn't trigger the rule "change 'me' to 'I' when it is a subject of a verb". This seems to be because the brain has already "clumped" the complex subject together into a single unit; it is no longer "Joe" and "me", but rather the single unit "Joe-and-me". The nominative-case-changing-rule in the brain, notices that "Joe-and-me" isn't "me", and thus doesn't make the change.

Of course, we have been taught since infancy that it should be "Joe and I"; but it seems, in English at least, this is a relatively difficult change to make. Probably this difficulty arises at least in part from the fact that nouns don't make any change, and two of the personal pronouns, "you" and "it", also don't change. So the English speaking brain does not have a broad, general "change-to-nominative-case-rule" based broadly on grammatical but rather a more limited "change-(some)-pronouns-to-nominative-case-rule" based more of the postion of the pronoun raltive to the verb which make it much easier to miss these pronouns when they have already been lumped together into a nounlike phrase that will become the subject.

So, back to overcorrection. Our Enlish-speaking brains have difficulty "seeing" the embedded pronouns in a nounlike phrase, so somehow the rule, or more likely a separate rule, gets set up along the lines of "in cases of complex subject containing "me", change the "me" to "I". Since we don't have in English a general subject/object change in nouns, the link to the grammatical function of the nounlike phrase is hard to see, to the rule gets applies too generally, and we see things like "the power of you and I" .

I do need to note an exception to this general rule that the objective is more basic than he nominative, and that is the word "who". "Who", the nominative form with the objective form "whom", seems to be the more basic form, increasing driving out the objective form; "whom" seems in many cases to be somewhat pedantic. The difference seems to be the use of the word, is is more frequently uses in the nominative case ("Who went?") that the objective ("whom did you say ...?) Also, note the apparent "trigger" to the rule to generate the nominative case above; it seems to be something along the lines of "stands alone before a verb". Since he grammar of "who"/"whom" is different from that of the other pronouns, "whom" frequently stands directly before a verb: "Whom did you say ..."). Further complicated by "you" and "it, which do not change form between nominative and objective usage, and thus prbobly weaken the overal strength of the rules that would make the change. All of which combine to produce "who" rather that "whom" as the basic form, with a rule to change "who" to "whom" unlike the other pronouns.

The South and West in U.S. presidential elections

In the last 20 years there have been 5 presidential elections. In two of these elections, those in 1992 and 1996, Ross Perot was a candidate who did much better than the average third party candidate, so we have had 12 "major" candidates: Clinton-1992, Bush-1992, Perot-1992, Clinton-1996, Dole-1996, Perot-1996, Bush-2000, Gore-2000, Bush-2004, Kerry-2004, Obama-2008 and McCain-2008. If we take the election percentages for each of the 50 states (leaving out the District of Columbia because it is such an outlier that it confuses the results), and run a component analysis on them, the results shows two significant components. Plotting the component scores for the various states, we get this:


The 10 states shown in red are, from the bottom up: MS, TN, AL, SC, LA, AR, GA, NC, VA, and KY.

The 13 state shown in green are, from left to right: UT, ID, WY, NE, AK, OK, KS, ND, SD, MT, IN, AZ, and CO.

The purple state between the two, at about [-.8, -.3], is TX

The first component (the X-axis) generally orders the states from the red republican-leaning states (UT, ID, WY) at one end to the blue democratic leaning states at the other end; the five stats with the highest first component scores are MA, RI, NY, HI, and VT.

The second component separates the southern states from the rest of the states; the five most "unsouthern states" are ME, MT, ID, AK, and VT.

So (roughtly) the lower left quadrent are the southern states, and the upper left are what may be called the "interior west" (with Alaska fitting in with these states).

TX, being a state the borders the south and the west (there is a saying to the effect that the border between the south and the west runs between Dallas and Fort Worth), unsuprisingly lies between the two groups.

What we can notice is that the southern states show up as only moderately republican (taken as the first component), eight of the thirteen "western" states (UT, ID, WY, NE, AK, OK, KS, and ND) are are farther to the left than the leftmost southern state (AL). Additionally, the second component absolutely separates them. Placing Texas in either group, considereing it either a southern or as a western state, does not change this; the two groups would still not overlap in the second component.

While there doesn't seem to be any obvious groupings of the states to the right of the scatter chart, I find it somewhat striking how two groups do separate out so distinctly.

The verb "read"

In the above posts, I have written about some of my ideas about a case grammar approach to the grammar of verbs discussed by Scott DeLancey and the University of Oregon. In one of the papers, What MIGHT be innate: Perceptual structure in linguistic structure, DeLancey makes the comment that under this approach "at points the semantic plausibility gets stretched a bit", and sets forth the interpretation

I	read	The King In Yellow.
AG	LOC	THEME

That is, DeLancey presents "read" as a loc-incorporating verb, with "I" being the agent that CAUSEs the Theme "The King in Yellow" to GOTO the Loc, "read", or "being read". This takes "read" to be analogous to "break" verbs, in which Agent CAUSE Theme GOTO Loc.

I am going to disagree with him here, arguing that "read" is more like a "hit" verb, incorporating the theme, roughly "reading"; leaving what is being read, "The king in Yellow", as the Loc to which the reading is directed. That is, I am arguing for

I	read	The King In Yellow.
AG	THEME	LOC

The verb "break", incorporates the resultant Loc, with the Theme going to that state. Note the difference between Joe broke the dish and The dish broke. In both the Theme (the dish) ends up in the resultant state (being broken). The verb "hit", which incorporates the Theme (one or more instances of hitting), which is directed to the Loc (what is being hit). So we have Jane hit the ball (Jane delivered a hit to the ball), but *The ball hit is impermissible. "Read", like "hit" and unlike "break", does not allow this alternation, *"'The King in Yellow' read" is not acceptable.

Also, we can say "Jane gave the ball a quick hit", which makes more explicit the transfer of an instance of hitting to the ball, but not *"Joe gave the plate a break" because, according to the theory, an instance of breaking is not being delivered to the plate as in a "hit" verb, but rather the plate is being delivered to the state of being broken. While perhaps a trifle non-standard, it is perfectly understandable to say "I gave the report a quick read", that is, as with "hit" but not with "break", which suggests that the event of reading is being delivered to the report rather that the report being given to the state of being read.

Using the three levels of meaning in the previous post, "read" acts like "kick" (and "hit"):

• Abigail read (Abigail AT (act of) read (-ing))


• Abigail read "Utopia" (Abigail CAUSE (act of) read (-ing) GOTO "Utopia")


• Abigail read "Utopia" to Fred (Abigail CAUSE "Utopia" GOTO Fred)

Does this analysis of "read" as a theme-incorporating verb address the problem noted by DeLancey of the "stretching of the semantic plausibility"? I believe it does. I take "read" to be what might be called an "attention directing verb", in which we claim that the attention of the agent is being directed to or toward an object. This is a difference between "hear" and "listen" (I heard it but I didn't listen to it -- the sound reached my ears, but I didn’t direct my attention to it) and between "see" and "look at" (I saw it but didn't really look at it). If "read" is such an attention directing verb, then "Sarah read the book" means something along the line of "Sarah directed her reading-attention to the book". Which seem to me to have semantic plausibility that treating "read" as a loc-incorporating verb lacks.

Kicking as a three-level verb

The simple "atoms" of verbs set forth in previous posts, with a THEME, LOC, and (perhaps) a CAUSE, are somewhat too simple to stand alone in analyzing most verbs.

Delancey seems to take it that "kick" is a "surface contact verb" and that she kicked the ball can be analyzed something like she CAUSE kick GOTO ball. ¹

He also seems to take it that she kicked him the ball is something like she CAUSE ball GOTO him.²

In the second sentence, while it is certainly correct that she kicked him the ball normally describes an event which can be described as she CAUSE ball GOTO him, that analysis lacks any indication of how she CAUSEd the ball GOTO him, while the sentence does tell us this - she kicked it to him. Note that the conversation She threw him the ball - No, she kicked it, makes perfect sense.

Also, this conversation seems to sugggest, although perhaps less obviously, there is a level of the kind of activity involved - kicking. She was involved in the act of kicking, not of, for example, throwing.

So it seems that the sentence She Kicked him the ball seems to analyze into three "layers": (1) Her act of kicking, (2) the causing of a kick to go to the ball, and (3) the causing of the ball to go to him.

Each of the first two act as a kind of "context" for those following it. She acted to kick (moved the leg in a striking motion), which caused a kick (a strike with the foot) to contact the ball, which in turn, caused the ball to go to him. Thus it was her act of kicking, and not merely "she", that CAUSEd the kick to GOTO the ball. Note that each level can succeed while those below can fail "she kicked (but missed the ball", and she kicked the ball (but it went wide). Thus each layer seems to be the cause of the later phrase. We thus now have an analysis of "she kicked him the ball" as

[1]: she AT (act of) kick (-ing)
[2]: [1] CAUSE kick (= strike with foot) GOT0 ball
[3]: [2] CAUSE ball GOT him

Note that being engaged it an activity is metaphorically being located AT the activity.

This analysis seems also to bring out the possible sentence rather naturally. She kicked the ball is [1] and [2]; while she kicked (but missed) is [1] alone.


¹See the discussion of surface contact verbs, including "kick", in LSA Summer Institute, UC Santa Barbara, 2001, Lecture 3 and What an Innatist Argument Should Look Like

²Delancey doesn't explicitly discuss "kick" as a ditransitive verb, but this observation seems to fairly follow for the discussion of ditransitive verbs in LSA Summer Institute, UC Santa Barbara, 2001, Lecture 3 and What an Innatist Argument Should Look Like