A contrastive analysis of dar ‘give’ in English and Brazilian Portuguese: semantic-syntactic relationships and implications for L2 instruction

In recent years, research on multiword lexical units has influenced second language acquisition research, but little work has been done on light verbs, especially comparing the use of light verbs in English and Brazilian Portuguese. This paper presents a comparative analysis of the syntactic and semantic aspects of dar and ‘give’ through the semantic continuum, event type, denominal verbs and incorporation. This study finds that distinct and varied semantic uses of light verbs present a unique challenge to second language learners in terms of both their understanding and their production. Furthermore, this study analyzes the semantic-syntactic interrelationships and suggests implications for teaching English and Portuguese light verbs to second language learners.


INTRoduCTIoN
A trend in recent years of growing interest in multiword lexical units has certainly influenced second language acquisition research (Folse, 2004). While this interest has increased studies on phrasal verbs, consisting of a verb and particle (ie., ate up), in recent studies including spoken corpus data (Lee, 2015) and avoidance (Liao & Fukuya, 2004), for example, not all multiword lexical units have received similar attention. Another type of multiword verbal construction is the light verb, which consists of a verb and noun (ie, give a wash, take a walk, have a rest). The term itself, coined by Jerspersen (1965), refers to the way in which the first verb in the light verb construction takes on less lexical meaning than a typical action verb. Live (1973) describes the first verbs in the light verb construction as "almost completely devoid of lexical meaning but [embodying] the associated grammatical information, being the bearer of the inflectional endings (thus indicating tense, number and person)" (Live, 1973, p. 31). This suggests that light verbs are by definition distinct in terms of their semantic composition. Through an analysis of argument structure, Pederneira (2014) explains that the syntactic structure gives rise to the distinct meanings of light verb constructions.
Unlike other multiword lexical units, the light verb has not been given equal treatment in the literature. In particular, there is a notable void in its application to language learning and pedagogy. The present paper, then, presents a unique contribution to the literature by presenting a contrastive analysis of the Brazilian Portuguese light verb 'dar' and its English counterpart 'give', considering how their semantic and syntactic composition pose linguistic challenges to second language learners (L2 learners). Additionally, a close study of this complex, frequently used verb demonstrates the inextricable quality of semantic-syntactic relationships (Costa, 2004).
The paper first reviews the concept of a light verb and its significance as a challenge for L2 learners. This is followed by a discussion of the semantic role of light verbs, with specific examples from English and Brazilian Portuguese. From there, the paper considers how give differs from dar on a structural/syntactic level. Finally, the paper highlights the interrelated quality of semantics and syntax in light verbs, leading to a discussion of possible pedagogical implications in the language learning classroom.

LITERATuRE REvIEw
Light verb constructions (LVC) are characterized by expressions in which "the main semantic content of the predicate is provided not by the verb, but the nominal complement" (Kearns, 1998(Kearns, /2002. These are traditionally depicted by V + N constructions; for example, "Jane gave the car a wash" means "Jane washed the car." However, there is significant debate about the specific qualifications of a light verb and the characteristics of their predicates. After all, the semantic difference between "giving a bath and "taking a bath" demonstrates that light verbs are not completely devoid of meaning. In essence, the central question could be framed as: Where does the meaning originate? Is it a syntactic or lexical/semantics contribution?

Original Article
O'Donnell Christoffersen, K. | A contrastive analysis of dar 'give' ... Kearns (1998Kearns ( /2002 delimits light verbs into two distinct categories: true light verbs (TLV) and vague action verbs (VAV). He argues that the vague action verbs (VAV) aren't as light, but have simply been "bleached." Di Sciullo and Rosen (1990), on the other hand, extend their analysis of light verbs to restructuring verbs which are "fully specified in meaning, but … actually light with respect to the argument structure" (p. 109). One example would be come in constructions such as "He came to realize". Linguists also disagree on the characterization of light verbs according to structure and behavior, which typically align with three different approaches. First, there are those who argue that light verbs lack lexical meaning and instead act as functional heads which bear tense and agreement features (Gross, 1981;Cattell, 1984;Grimshaw & Mester, 1988). The second camp posits that light verbs are auxiliaries with aspectual features (Hook, 1974;Abiellé, Godard & Sag, 1998). The last approach, and the perspective taken up in this paper, considers light verbs as a specific subclass of verbs that play a significant role in predication (Rosen, 1990;Alsina, 1996;Butt & Geuder, 2001;Butt, 2003Butt, , 2010Samek-Lodovici, 2003;Duarte, 2009;Pederneira, 2014). Specifically, this analysis follows Butt's (2010) definition light verbs as complex predicates which include two or more elements which contribute to joint predication, are form identical to a full verb, and monoclausal.
Given the challenges linguists face when classifying light verbs, it is no surprise that light verbs also provide a challenge for second language (L2) learners. Some of the difficulties involve cross-linguistic variation in word choice, use in idiomatic expressions, ambiguity, and frequency. For instance, learners may question why in English it is possible to "get someone frustrated," but in Spanish someone may be "put frustrated" (pone frustrado). Similarly, in English it is possible to "make a turn," while in Brazilian Portuguese it would be "give a turn" (dar uma volta). These types of expressions are typically very confusing to L2 learners, and this is largely due to the fact that light verbs are among the least uniform words in terms of their meaning (Melamed, 1997). This variable meaning is especially problematic since many of the most common lexical verbs in the English language are often used in LVCs, such as 'get ', 'make', 'take', 'want', 'give' (Biber et al., 2002).
The frequency, semantic complexity, and unique syntax of dar 'give' make it an ideal candidate for a cross-linguistic comparison of light verbs. While Biber et. al. (2002) ranked 'give' as one of the most common word in the English language, Melamed (1997) demonstrated that forms of 'give' topped the list of verbs with the highest entropy, or most ambiguity, in translation in an analysis of parallel text corpora (p. 41). This ambiguity is certainly related to the verb's various functional subtypes. According to Live (1973), two of these subtypes are a) an instance of exertion or spontaneous response, i.e., give a grin, or b) resembling the dative or indirect object construction as in give her a call (p. 32, 35). These functions are related to the structures V+NP in either the noun (give a wash) or gerund form (give a cleaning). Some linguists such as Jespersen (1965)

SEmANTIC ANALySIS of dar 'gIvE'
The wording in the term "light verb" suggests that this functional category is set apart in terms of its semantic value, although as mentioned in the previous section the exact "semantic weight" is under debate. An alternative to defining a "weight limit" for the light verb category is taking the perspective of a semantic continuum along which verbs are placed according to their relative "abstractness" or "lightness" (Butt & Geuder, 2001;Newman, 1996). This follows from current understandings that light verbs are extensions of prototypical "full" verbs (Butt, 2010;Pederneira, 2015). Based on this reasoning, the semantic continuum provides a useful framework for an analysis of communicative function, ambiguity, event type, and causative constructions.

give: The semantic continuum
The light verb "give" is closely examined by Newman (1996) and further explored by Butt & Geuder (2001) who proposed an organization of ditransitive verbs along a continuum moving from "full" (concrete) to more "light" (abstract) meanings. Figure 1 visually represents the placement of Butt & Geuder's (2001) examples along the proposed continuum (1, 2b, 3, 4) including their examples of V+NP constructions (5a, 5b); however, Figure 1 extends their analysis with the addition of monotransitives (2a) and phrasal verbs (5c). figure 1: Semantic continuum of 'give', modified from Butt & Geuder (2001) Furthermore, as illustrated in (1), the most concrete realization of "give" may be explained as a "change in location accompanied by a change in possession" (Butt & Geuder, p. 340, 2001). In (1), when the ball is given, the possession changes from the subject of the sentence to "him," and there is an inherent change in location. As the prototypical understanding of "give," this is how most people will define the term regardless of the context of its usage.
(1) give him the ball give the dog a bone give the customer a recipe MEANING: change in possession causes a resulting change in location (1) give him the ball (2) a. give the children the inheritance. b. give light/milk/ heat.
(3) give advice (4) give someone emotional support (5) a. give the car a wash b. give someone a kiss c. give in
(2) a. Tom gave the children their inheritance. b. Cows give milk. MEANING: change in possession causes a result Continuing onward, (2a) appears to differ only slightly, in that what is being given isn't an actual physical object called an "inheritance" but rather the "rights to an inheritance" (Butt & Geuder, 2001, p. 341) which cannot actually change location. So, we might describe this example as a change in possession which causes some kind of unspecified result. In this case the result is that the children have the rights of the inheritance in their name (change in possession) but those rights did not move to a different location as opposed to the ball, the bone or the recipe in the previous example (1). Monotransitive verbs, like (2b) also fall under this level of meaning. In the sentence Cows give milk (2b), there is a change in the possession of the milk and a result from this change, but note that the location of the milk is not specified. This is because (2b) is relating the fact that giving milk is something that cows do, which can be compared to its ditransitive counterpart; i.e., The cow gave milk to its calf one day.
(3) give advice give someone the right to do something give someone information MEANING: change causes a result In (3), the meaning starts is more abstract, because "advice" and "rights" cannot act as physical property. While these readings of give do correlate with the notion of possession in an abstract sense (possession of wisdom, of freedom, of knowledge), there is a vague sort of relationship between the verb give and the NP. Once again, the focus is on the result; someone is gaining advice, rights or information.
(4) give someone emotional support give someone one's regards MEANING: exertion causes a result One step further and the result is even more of a focus in the sentence. Although the examples in (4) demonstrate results which may be beneficial to the recipient, the exertion is more certain than the outcomes of the result. To provide a beneficial result requires a degree of control. To use the example in (4), Lee might give Amy emotional support, but it is possible that Amy doesn't recognize that support. In such a case, although Lee exerted the effort, it might not actually change anything, at least in the expected or anticipated way.  (5). Although all demonstrate some exertion of an effect on the recipient, they differ in terms of movement-based and non-movement-based exertion. In giving someone a kiss, push, or punch (5b), there is an active form of movement involved. Here the visible movement might at least be interpreted as a "transfer," and in that way relate to the prototypical "full" verb give. Conversely, all that is left of the transaction in the phrasal verb give in (5c) is an application of an action to an entity, denoting a result only within the structure of the LVC itself. For instance, when someone says, "I gave in and went to the movies," the structure gave in only represents the resultant state of the speaker surrendering their own desire, while the rest of the sentence suggests what caused it.
The prototypical meaning of the word give is consequentially stretched from the direct transfer of location based on a change in possession as in giving him the ball all the way to the vague resultant state of giving up. In order to further understand the wide-ranging functional meaning of the verb give, we will distinguish these levels of abstraction based on individual semantic features relating to aspect.

give: Event Type
Although "aspect" originally referred to such grammaticized viewpoints (perfect and imperfect), it has been extended to encompass the relationship between certain viewpoints and situation types (Smith, 1997). Often referred to as internal event structure or 'aktionsart', we will use the term "event type" in concordance with Vendler's Aspectual Classes (1967). To further account for the relationship between aspect and meaning, Vendler states that, "The fact that verbs have tenses indicates that considerations involving the concept of time are relevant for their use" (Vendler, 1995, p. 143). However, he emphasizes the fact that this "time" that he refers to involves more than distinguishing past, present or future. The following examples based on Vendler's explanation may help to clarify this point.
Imagine that someone asks you the question, "What are you doing?" You might answer, "I am writing (working, studying, etc.)" but not "I am knowing (loving, recognizing, etc.)". This distinction illustrates that writing, working, studying, etc. are "processes going on in time, or successive phases following one another in time" (Vendler, 1995, p. 144). This is not so for verbs like know, love, recognize, etc. You may "know Spanish," but that does not mean that Spanish is a process that is going on right now. From this example, we can see that the difference in the use of these verbs demonstrates the notion of time inherent in their meanings. It is important to note that this may not be true for many verbs and is likely a contextual restriction rather than a syntactic/aspectual one.
It is on the basis of this relationship that we use Vendler's Aspectual Classes to more clearly analyze the meanings from the previously discussed continuum. In analyzing specific semantic features related to time, we will better understand the overall use of the verb give in different situations or event types. The three semantic features that will be examined here include punctuality (occurring instantaneously), telicity (having an inherent

Original Article
O'Donnell Christoffersen, K. | A contrastive analysis of dar 'give' ... endpoint), and dynamicity (describing an action). Such features provide a means for separating verbs into classes of: states, activities, accomplishments and achievements. It is important to note that Role and Reference Grammar is the grammatical theoretical framework that further expounded upon this classification system, describing the so-called "logical structure" of the predicate class (Foley & Van Valin, 1984;Van Valin, 1990).
The following chart displays how the different levels of verb "give" proposed by Butt & Geuder (2001) correspond to Vendler's Aspectual Classes (1967). The test questions, listed under the Aspectual categories were developed by Shirai & Anderson (1995). Section 3.2.4 uses a test for in/during/after which parallels tests of "for" and "in" used by Borer (2004Borer ( , 2005aBorer ( , 2005b See also Smith (1991).

Linguistic Tests
State Activity Accomplishment Achievement Punctual  (1) give him the ball (3) give advice (4) give someone emotional support (5b) give the car a wash (2a) give the children their inheritance (5a) give someone a kiss (5c) give in

State
As we remember, the prototypical meaning of give involves some sort of change of state; therefore, it is not surprising that there would be no stative reading. State readings answer the question: Does it have a habitual interpretation in the present tense? Verbs like see and hate do not; i.e., "I always see" or "He always hates". (Note that this is possible if a complement is added as in "I always see you at work". But it is not grammatical without the complement.) In terms of semantic features, a state lacks dynamic, telic, and punctual tendencies. In other words, "I see" is not active, doesn't have a determined endpoint and isn't instantaneous. However, even the least dynamic of all give readings is active in some way. This is related to commonality among the aforementioned meanings: a result from a change.

Activity
Monotransitive verbs such as "give milk/light/heat" are classified as activities, suggesting that their meanings encompass a dynamic aspect. Although "Cow gives milk" might not seem very active at first, effort is being applied in the giving of milk. Using the linguistic tests set forth by Shirai & Anderson, the fact that Cows give milk has a habitual interpretation in the present tense classifies it as an activity. Additionally, it can be likened to other verbs

Original Article
O'Donnell Christoffersen, K. | A contrastive analysis of dar 'give' ... like run, sing, play and dance that are constant in its quality throughout the duration of the activity. At every moment that the cows give milk, the giving is the same. It also has an arbitrary endpoint that is not built into the activity itself. In addition to these features specified by Vendler, we should note that there is no specified accomplishment or achieving inherent in the sentence Cows give milk, since here it is talking about an activity that cows usually take part in, not a specific instance. Compare with its ditransitive counterpart; i.e., The cow gave milk to its calf this morning.

accomplishments
In addition to being dynamic, verbs in the accomplishment category are also telic, or have a determined endpoint. However, they are different from accomplishments since they have some duration. For example, give him the ball is a dynamic action, and it also has a determined end point, when the ball is given. But, the fact that there is some duration of time involved in giving him the ball makes it different from an achievement. Giving advice (3) and giving emotional support (4) work similarly, since they are active events that require a certain time but have an endpoint. And while giving the car a wash (5a) may appear to be more complicated, the aspect of the event follows the same pattern: it is an action, it has an end, and it requires some amount of time.

achievements
The difference between the event type accomplishment and that of achievement is that achievements are punctual (instantaneous). Accomplishments often result in products; examples include make a chair, build a house, and run a mile. The products in the previous examples are: the rights of ownership given to the kids (2a), the kiss given to someone (5b), and the surrender to someone (5c). In other examples of accomplishments, the product may not be clear, such as 'climb a mountain' or 'push the car'. In these instances, it is beneficial to employ linguistic tests for achievements/accomplishments outlined by (Shirai & Andersen, 1995). Borer (2004Borer ( , 2005aBorer ( , 2005b) has a similar test using "in" and "for", which parallels these in/during/after tests. a) 'X Ved in Y time,' is not the same as 'X was Ving during that time. ' "He gave the children the inheritance in 10 minutes" is not the same as "He was giving them the inheritance during 10 minutes." b) There is no ambiguity with 'almost.' He almost gave the children the inheritance.
(No difference between he almost finished or almost started.) c) 'X will VP in Y time' is not the same as 'X will VP after Y time.' "He will give the children the inheritance in 10 minutes" is not the same as "He will give the children the inheritance after 10 minutes."

Variation
It is interesting to point out that while levels (1), (3), (4), and (5a) from the abstractness continuum can be classified as accomplishments, (2a), (5b), and (5c) are determined to be achievements. This varied alternation suggests that punctual (instantaneous) and unpunctual features do not influence a verb's perceived level of abstractness. In other words, perceptions of abstractness or concreteness are not directly related to whether that event happens immediately. Additionally, the event type for monotransitive verbs (2b) is activity, while prototypical verbs (1) are accomplishments. This nonlinear pattern again suggests that telicity (having a clear end point) does not directly determine an event's place along the continuum. One must also not discount the possibility that aspect is rooted in syntax and not lexical/semantics as it has so often been analyzed.

dar: The Semantic Continuum
Now, since we are familiar with differing levels of abstraction on the semantic continuum as well as categories of event types, it is in our interest to compare this common light verb "give" with its Portuguese counterpart "dar." (1) dar a bola a ele [give-the-ballto-him] (2) a. dar aos meninos a herança [give-to-the-children- figure 2: Semantic continuum of 'dar' While you may notice some differences in terms of word order and the 3 rd person direct object complement (lhe), these are syntactic issues that will be covered later in this paper. In terms of the semantic meaning and levels of abstraction, the verb lines up in both languages in terms of the meanings designated by categories (1) through (5). Here is a quick overview of the respective meanings: (1) change in possession causes a resulting change in location (2) change in possession causes a result (3) change causes a result (4) exertion causes a beneficial result (5a) movement-based exertion causes a result (5b) non-movement-based exertion causes a result (5c) result from a change (unspecified/unmentioned) Also, notice that (5c) is no longer "give up" but the multi-word phrase dar conta de. This literally means "to give account of" and is usually translated as "to realize". This expression adheres to the meaning in (5c), since when you realize something it implies a change that causes the resulting realization. Still, the similarities suggested by this analysis are really quite deceptive.

dar: Event Type
A further point of comparison for 'dar' and give is Vendler's Aspectual Class. The following chart displays how the different levels of verb "dar" correspond to Vendler's Aspectual Classes (1967) and the questions developed by Shirai & Anderson (1995).

Linguistic Tests State Activity Accomplishment Achievement
Punctual

State, activity, accomplishment, achievement
The state and activity readings for dar are the same for both English and Portuguese, but as can be noted by the bold-face type there are some different realizations of (5a) and (5b) that wouldn't require a distinction in English. How is "giving a wash to the car" different from "giving a screw to the furniture," and does that even work in English? For (5b), we are prompted to ask why "giving a knifing" is in the category of achievements, when in English it would remain in the accomplishment category. Furthermore, how is "giving a knifing" distinct from "giving a wash/screw"? The awkward sounding phrases show that dar is functioning differently than give in (5a) and (5b). Additionally, dar as it is used in (5c) is a multi-word phrase taking the place of verbal phrases from the English examples. How is it similar or different? It is clear that the semantic continuum and event type schema are no longer sufficient to answer our questions about the cross-linguistic usage of these verbs. In fact, perhaps there is no lexical semantics here but rather a semantic interpretation which originates from argument/syntactic structure.

A SyNTACTIC ANALySIS of dAR/gIvE
One of the most noticeable qualities of the analysis of dar in terms of Aspectual Class is that there are certain word choices within the "give" constructions that are preferable to Portuguese speakers but awkward to English speakers. But why would this be? What determines this distinction? What kind of systematic pattern exists to determine their language specific uses?

Analysis of denominal verbs
You might have already noticed that the predicate arguments in the "give" LVC may be rephrased as verbs. For instance, "Susan gave the car a wash" could be conflated to "Susan washed the car." "Wash" here would be considered a "denominal verb", since it has been projected from the predicate into the verb phrase. Clark & Clark (1979) devised an extensive catalogue of these denominal verbs, and the following comparison explores verbs that may be derived from give according to their locatum and instrument categories. The examples move through three types of constructions 1) give the car a washing, 2) give the car a wash, 3) wash the car. Looking briefly over this analysis, the first thing that might stand out is that "locatum" verbs (6) of giving to an animal or a person work similarly in English and Portuguese. Here we could also insert our example from (5b); i.e., give the boy a kissing, give the boy a kiss, kiss the boy. In terms of instruments (7), the "clean" verbs seem very flexible in English, while Portuguese resists incorporation of certain cleaning instruments into the verb position, especially personal care items such as shampoo.
(Other examples include soap-*sabonete-ar and hairspray-*laque-ar.) On the other hand, "hit" verbs seem more open to a variety of instruments in Portuguese, where the ending -ada suggests hit with. This explains the alternate (5b) example of

Original Article
O'Donnell Christoffersen, K. | A contrastive analysis of dar 'give' ... knifing/facada. While most of the predicate structures in English are related to actual things (nouns), the words in parenthesis in the Portuguese side of the chart show that these verbs are not derived directly from the instrument used in the process. This accounts for the alternative (5a) example dar uma passada or give an ironing. While the Portuguese word for iron is ferro, the word for ironing is passar. Through this difference, it is possible to note that the verb (not the noun) is used for -ada constructions. Scher (2004Scher ( , 2006 analyzed -ada constructions in detail, and this work will be further explained below. So the question is still, but how? Why are these different trends possible? Hale & Keyser (1993) extended Clark & Clark's work on denominal verbs in a syntactic view of lexical argument structure. According to this perspective, "each lexical head projects its category to a phrasal level and determines within that projection an unambiguous system of structural relations holding between that head, its categorical projections, and its arguments" (p. 53, 1993). The following example may clarify how this projection takes place within the structure of the syntax. Tree diagrams (8) and (9) show the "incorporation" of the noun "a wash" into the verb "wash" in the sentence. Following Hale & Keyser (1993), the HAVE/POSSESS/CHANGE and CAUSE are both lexical/semantic and

Original Article
O'Donnell Christoffersen, K. | A contrastive analysis of dar 'give ' ... syntactic. These categories, or "flavors of v" (Folli & Harley, 2005), project a system of structural relationships. The HAVE/POSSESS/CHANGE which forms "give" together with CAUSE, correlates directly with previously noted meanings (1) through (5) from most concrete to most abstract, remembering that the causal relationship is what remained the same. This diagram clarifies how the LVC containing "give" really represent an underlying causal relationship. He gives the car a wash represents the fact that he is causing a change such that the car becomes washed.
According to Scher (2006), the possible incorporation of a predicate structure as well as its aspectual meaning provide clues to its usage. This essentially lexical semantic approach which combines aspect and incorporation will explain the additional examples of (5a) and (5b) in the achievement and accomplishment categories respectively.

dar: Achievements
(5a) dar uma limpada no carro [to give-a-cleaning-to the-car] Scher determines three types of constructions using the dar -ada construction. The first of these is in the first example of (5a). Here the form limpada is derived directly from the root verb limpar. As such, the aspect carried by the verb would remain with that construction. That means that the tree diagram of such a sentence would look like this: (9) In this case, the aspect of the word would be an eventuality (an accomplishment) usually with a diminuitive reading. A diminuitive reading can be described as anything done in an incomplete, quick or careless manner. For instance, you may say that you "gave it a reading," but there's still quite a lot left until you can say that you "read it." While the denominal verb structures (read) are related to LVC (give a reading), they do not entail one another completely in terms of meaning. That is why light verbs are only light and not completely semantically empty.
In Portuguese, these roots as verbs are usually accompanied by a prefix or affix in order to differentiate them from the nouns. In fact, the verb is differentiated from a noun with a verbalizer or nominalizer, such asear in esfaquear. For example, the root "fac" would make this transition: fac/esfaquear/esfaqueada. The roots in (10) are all directly related to verbs in Portuguese; however, this is not the case of their English counterparts. According to this explanation, the reason you can't give a "knifing" or a "fisting" or "currenting" but you can give a "stoning", "heading", "hammering" or "rowing" is that the prior terms are not verbs in English

Original Article
O'Donnell Christoffersen, K. | A contrastive analysis of dar 'give ' ... while the latter are. It also determines why these verbs are achievements, or diminuitive incomplete actions. Picture the difference between, "The team rowed up the stream," and "The team gave a row up the stream." (5a) dar uma parafusada nos móveis [to give-3pDO-a screwing-to the-furniture] The distinction between the previous (5a) example and this one are slight. In fact, the difference isn't apparent at the surface level. Instead, we have to use Scher's analysis to go deeper. (10) In this case, the main difference is that the root of the LVC is derived from a corresponding noun. Although the aspect or diminuitive reading remains the same, the difference is what is actually packed into the verb form. This can be demonstrated by Kiparksy's (1982) analysis of "hammer" and "chain" in English. He notes that you can "chain a prisoner with a chain" and not with anything else, because the meaning of "chain" is "to tie up with a chain." In comparison, you could "hammer a nail with your shoe," since the meaning of "hammer" is not to "hit with a hammer" but "to hit with a flat surface." Incorporating aspect into this perspective, Kiparsky's "hammer" and "chain" examples would look like this: (11) (Scher, 2006, p. 36) This wouldn't seem to mean a great deal to our present analysis except for the fact that this means something about how the words can be used. An L2 learner will need to learn how verbalizers and nominalizers work within the framework of the language they are learning, especially given the significant differences between light verbs in Portuguese and English. There is really no difference between the achievements in these two examples, except for the noticeable one. The first example ends in a noun, but the second example has the -ada ending, which English speakers associate with achievements but may be an Event. For example, if I were to say, "I gave the theif a knifing" in English, it doesn't seem countable like its previous counterpart. Notice that you may say, "I gave him three kisses," but not "I gave him three knifings." However, the underlying structure demonstrates what is actually going on here.

dar: Achievements
(12) (Scher, 2006, p. 37) Remember that in Portuguese, the verb related to the noun "knife" is "esfaquear." This is due to the -ear ending in the casa of faca. Therefore, when you give an "esfaqueada," it has the accomplishment reading. However, when you give a "facada" to someone, it has a noun reading. That would make it act like other countable nouns: kiss, push, nudge, etc. Additionally, this tree diagram loses the Asp and consequentially there is no diminuitive reading. Our previous example demonstrated that there is an incomplete quality about "giving a reading," suggesting it is an activity which was not finished. This is not so about "giving a kiss."

PEdAgogICAL ImPLICATIoNS
From the aforementioned analysis, we can see that the light verb dar would challenge second language learners in terms of both their understanding and their production of correct utterances. Yet this is not just theory, as critics of constrastive analysis often claim. The aforementioned research by Biber et. al. (2002) and Melamed (1997) confirm that give is problematic for second language learners. An additional study by Specia (2005) analyzed 200 instances of the verb give in the tagged corpus Compara and found the accuracy of the translations to be 91%.
While the students do need to learn how to construct the actual light verb phrases in using the infinitive or gerund form instead of the -ada, this is a minor concern. The trickiest aspects of dar constructions lie in Scher's deep structure. For instance, in the categorical chart depicting differences, we noticed that Portuguese-speakers could say Dei uma sapatada no cachorro, literally "I gave a shoe-ing to the dog". This stems from the fact that shoe here is a noun that would fit into the (5b) I gave a kiss construction, making it an accomplishment. While in Portuguese, nouns such as shoe and knife may be put into dar constructions to give the reading "hit with a ____", this is not possible in English.

Question
Test Accomplishment 1 Accomplishment 2 Achievement I gave the clothes an ironing.
I gave the clothes an ironing is less complete than I ironed the clothes.
I gave her a kiss.
I gave her a kiss 3 times.
If students understand to recognize the roots as the actual objects (nouns) or the verbs (actions) that they are dealing with, then students will have the keys to understanding what they can do with language. More importantly, they will understand the wide range of ways in which meaning can be packed into words, and even though "give" is a light verb, it is weighty in terms of its importance.
Beyond explanations and examples, a large quantity of natural language input is crucial for L2 learners. Academic and textbook materials often are devoid of common phrasal verbs and light verb constructions, although these are very common in informal everday conversations. Waara (2004) points out that input is "substantially less" in a classroom situation than a natural language learning setting. Online immersive environments and video conversations with speakers of English/Portuguese are ways that teachers can use technology to provide additional and varied informal input for L2 learners.

CoNCLuSIoN
The present analysis of the syntactic and semantic aspects of the light verb dar 'give' in Brazilian Portuguese and English has served to highlight the importance of the view of light verbs as complex predicates with which include two or more elements which contribute to joint predication, are form identical to a full verb, and monoclausal (as in Butt, 2010). Additionally, it is clear that the semantic continuum and event type schema are insufficient alone without taking into account syntactic structure. In fact, the analysis seems to suggest that there may not be lexical semantics at play here but rather a semantic interpretation which originates from argument/syntactic structure. Finally, pedagogical implications for teaching light verbs to L2 learners include use of examples, explanations and charts (as in Shiria & Anderson, 1992) as well as plentiful natural informal language input. This could include the use of online immersive environments as video conversations with speakers of English/Portuguese.