The occurrence of null objects varies across languages. An illustrative grammatical difference between Portuguese and English comes from object drop. While definite and indefinite object drop are allowed in Portuguese (; Cyrino , ), English does not (commonly) allow object drop (; Allerton 1975; see ; ; Groefsema 1996 for exceptions), as shown in (1) and (2).
Other languages disallow definite object drop, but license indefinite object drop, like (European) Spanish (Campos , ; ), as shown in B’s answers in (3) .
The asymmetry in (3) has been explained in light of the pronominal system of Spanish. Since definite clitics are available in Spanish, it is necessary to use them to refer to definite antecedents, as in (3a) (i.e., definite object drop is not allowed). Additionally, since Spanish lacks partitive clitics, null objects can be used to refer to indefinite antecedents, as shown in (3b) (i.e., indefinite object drop is allowed). Languages displaying partitive clitics, like Catalan and Italian, must make use of them instead of null objects in cases like the one in (3b) (). The generalization that captures these facts appears in (4).
The generalization in (4) predicts the behavior of (European) Spanish in (3) when referring to (in)definite antecedents from object positions. However, a crucial challenge for the generalization in (4) is outlined in the following section: the existence of referential Definite Null Objects (henceforth, DNO) in Spanish Recipes (henceforth, SR).
Interestingly, there are certain contexts where definite object drop can be found in languages that do not commonly license it, such as English and (European) Spanish. Definite object drop was first noticed to be possible in English recipes (; ; ; ; ; ), as shown in (5a), user instructions (; ; ), as shown in (5b), and sports broadcasting (; ; ; ; ), as shown in (5c).
Examples like the ones in (5a-b) belong to a broader register: the instructional register, where the addressee is supposed to carry out an instruction. This paper only discusses examples such as (5a), from the recipe register. Crucially, examples such as (5a) can also be found in SRs. Note that the null objects in (6), from Spanish, receive definite, specific readings, and refer to equally definite, specific antecedents. Additionally, DNOs can alternate with clitics in SRs, as shown in cuécelo ‘boil it’ in (6). This alternation is also possible in English recipes, and does not imply different meanings. That is why it has been said to be dictated by stylistic conventions ().
Agrega [el huevo]i y cuéceloi. Retira Øi, refresca Øi, pela Øi y reserva Øi.
‘Add the egg and boil it. Remove, refresh, peel, and reserve it.’
The antecedents of DNOs are ingredients in SRs, and they are quite accessible, as each ingredient is specified in the recipe list (; ; ). Additionally, the ingredients are contextually present, thus the information about them is deictically obtained (in fact, cooking situations are evoked even when reading written recipes). This favors the prominence of the referent introduced by the antecedent, and makes it easier to retrieve the referents of DNOs. Specifically, DNOs in these cases are “contextually determined and the interpretation is guided by the nature of the recipe register” ().
The problem is (at least) twofold. The first problem is to explain why, if Spanish displays clitics for referring to definite antecedents, null objects are used in (6), and still the sentence is grammatical, as opposed to the sentence in (3a). In light of this problem, the first question is what grammatical conditions license DNOs in SRs. In this regard, the first claim in this paper is that imperatives (and, more broadly, directive utterances within instructional registers) explain the availability of DNOs in SRs. The second problem is that the DNOs-clitics alternation does not trigger a difference in meaning: the antecedents of null objects, the null objects themselves and the clitics are interpreted as definite and specific. The second basic question is then what analysis can predict the behavior of the ellipsis site and this alternation in SRs. In this regard, the second claim in this paper is that these similarities between DNOs and clitics, along with the clitic-DNO alternation, suggest that DNOs display the behavior of (null) pronominals.
As for methodology, the data were collected from three web pages: Cocina Abierta (see Appendix A), Le Cordon Bleu (see Appendix B), and Recetas MasterChef (see Appendix C), available on the internet. This methodological choice responds to the criticism in the literature on analyses of DNOs in English recipes that collected isolated sentences, thus not reflecting real recipes (). The corpus amounts to 15 recipes, 5 from each web page. Each recipe was selected randomly, and DNOs and clitics were manually coded and highlighted in bold. The number of sentences amounts to 441, where 143 occurrences of DNOs and 43 occurrences of clitics were coded (i.e., the occurrence of DNOs is more frequent than the occurrence of clitics). The selected texts can be found in the Appendix. Only the description of the preparation of the meal is reproduced (the sections preceding the recipes are not included, since they are not linguistically relevant for this study). Note that DNOs in written recipes might behave differently than DNOs in videos, where the referents are deictically available. Future research might shed light on these cases.
This paper is organized as follows. In Section 2, I review some analyses for DNOs in English recipes, such as those by wh-trace (2.1), NP-trace (2.2), nP (2.3) and article drop (2.4). I conclude that they cannot be extended to DNOs in SRs. In Section 3, I present some basic grammatical conditions and properties of DNOs in SRs. Specifically, this section is devoted to affectedness (3.1), directive utterances (3.2), syntactic representation (3.3) and DNOs without linguistic antecedents (3.4). In Section 4, I wrap up the ideas (4), and put forward an account for DNOs in SRs, which relies on null pronominals in object position (Section 5). Finally, I conclude with some observations (Section 6).
Given the lack of research on DNOs in SRs, this section discusses previous analyses for English recipes, and tries to test if any of these analyses can be extended to DNOs in SRs. The following sections present the topic drop analysis (2.1), the NP-trace analysis (2.2), the nP analysis (2.3) and the article drop analysis (2.4) are presented in the following sections (see ; Paul & Massam , for Malagasy; ).
According to Haegeman (, ), DNOs behave as topics linked to an empty category in the matrix clause. The empty category occurring in object position is the trace of an operator (Chomsky , ; ). Therefore, DNOs are topicalized in the left periphery, and subsequently deleted, as illustrated in (7).
Since the operation in (7) triggers movement, the general constraints on movement must be respected. One of them, the Condition on Extraction Domains, states that a phrase may be extracted out of a domain if that domain is properly governed (; ). According to Haegeman (, ), DNOs cannot be extracted from ‘strong’ islands (e.g., relative clauses), as shown in (8). This seems to confirm that DNOs are wh-words that have been moved from their base position.
?Lift [the chicken pieces]i, preserving the mixture where you have marinated Øi.
This analysis correctly explains that DNOs in recipes behave as topics and refer to prominent antecedents. However, it faces some challenges. First, “since wh-questions are implausible in recipes, it is impossible to see if the operator [...] is in complementary distribution with the question operator, a prediction made by the analysis” (). Hence, the proposal lacks predictive capability. Second, the verb selecting the DNO in (8) does not display an imperative form, as opposed to the example in (5a). This suggests that the grammatical properties of the sentence in (8) are closer to the ones of the sentence in (2a), from the ‘neutral’ register of English, which does not allow DNOs. In other words, finite clauses “might be a case of switching into the neutral register” (). The same ungrammaticality is expected in the Spanish counterpart of (8), as the Spanish ‘neutral’ register does not allow DNOs. Hence, the ungrammaticality of the sequence in (8) is not related to constraints on movement or island-sensitivity (i.e., DNOs are not instances of wh-traces).
Interestingly, as stated in the literature, DNOs in recipes only occur in sentences with null subjects. This condition is striking, because English is said to be a non-null-subject language (). However, null subjects can be found in English in coordinated and imperative sentences (Haegeman , , , ; Weir , ). Since verbs in recipes display imperative values, null subjects can be licensed. Crucially, a sentence with DNOs and null subjects, as in (9a), is allowed in recipes, but a sentence with DNOs and overt subjects, such as (9b), is not.
Therefore, according to Massam & Roberge (), the occurrence of DNOs in the recipe context depends on the occurrence of null subjects. This leads Massam & Roberge () to conclude that DNOs behave as NP-traces (i.e., traces bound by an element in A-position). In this account, the binder of the NP-trace is a null topic that occurs in subject position. This is said to predict the impossibility of DNOs occurring with overt subjects in the same sentence, as shown in (9b), since the subject position is occupied by an operator.
This account is not exempt from criticism. First, it is only possible on the assumption that imperative sentences with null subjects lack a syntactically projected subject before the operator moves onto the subject position. It is difficult to accept this approach for Spanish, a consistent null-subject language where null subjects are syntactically projected (; ). Second, it is unclear if the ungrammaticality of (9b) is strictly related to the operator being unable to occupy the subject position. As said above, English restricts the occurrence of null subjects to imperative sentences. Given that null subjects are the default with imperatives, the ‘marked’ grammatical form of imperative sentences is the one with overt subjects. Hence, overt subjects only occur under certain pragmatic conditions (i.e., when reinforcing the value of the imperative, and implying decisive orders). Since such pragmatic conditions are not met in recipes, the oddity of the sentence in (9b) may be due to this fact. Additionally, in (European) Spanish, imperative sentences with overt subjects, as the ones in (10), are the ‘marked’ grammatical forms, so their occurrence is restricted to certain contexts. One of them is that of contrast (; see ; ; on the pronominal system of Caribbean Spanish varieties).
#Bate Ø tú bien y cocina Ø tú durante 5’.
‘You beat it well and cook it for 5’.’
Since there is no contrast in (10) (because the referent of the subject is accessible), the sentence is strange. Hence, the (alleged) ungrammaticality of (9b) is unrelated to the occurrence of an operator in subject position (i.e., DNOs cannot be analyzed as NP-traces).
From a minimalist perspective, argues that DNOs in recipes are better analyzed as nP. Prior to defend this claim, Ruda () presupposes that DNOs are derived by ellipsis, in line with (part of) the studies on ellipsis, which state that null arguments raise as a result of a full phrase projected in syntax (; ; ). Recall that ellipsis sites must refer to antecedents (i.e., they must be anaphoric). However, the corpus in Ruda () contains examples of DNOs without linguistic antecedents. The antecedent of the DNO of spread is not linguistically represented in (11). Rather, “it is the resulting mixture which should be spread over the meringue rather than the yogurt, cream, sugar, and zest separately” (). This suggests that analyzing DNOs in terms of ellipsis is not workable.
Put the yogurt, cream, sugar and zest in a bowl; whip Ø to soft peaks. Spread Ø
evenly over the meringue.
Next, it is claimed in that DNOs display pronominal behavior in recipes, since DNOs and pronominals share definite readings, and DNOs can be replaced by pronominals. Hence, the syntactic representation of DNOs should be that of pronominal elements. Pronouns are projected as DPs in English (; , and contain a Num head, the lowest head in the projection of a pronoun being ., which contains gender (; ), as illustrated in (12). Recall that pronouns must realize the highest projection (i.e., DP), to be spelled-out ().
In line with ) on null objects in European Portuguese, analyzes DNOs in recipes as truncated pronouns projected as nP. If pronouns are only spelled-out when targeting DP, the DNO in (13), represented as nP, is not overtly realized as a pronoun. Hence, this proposal assumes that “the structure merged in the object position in the recipe register lacks both the D and the Num heads” ().
The crucial advantage of this proposal is that it captures the relation between DNOs and imperatives in recipes. It has been assumed in the literature that languages lacking DP projections lack TP projections (). Recall that imperatives are said to lack T (), and DNOs are said to behave as nP, not as DP (). Consequently, the correlation between the lack of T and the lack of D is captured.
However, the proposal faces some challenges. First, overt truncated pronouns do not occur in recipes, so it is not reasonable that only their null counterpart can occur. Second, DNOs in SRs display number features, since they trigger agreement with anaphoric expressions, as shown below (3.3). Thus, DNOs display number features in SRs (i.e., Num must be projected). Finally, DNOs are visible as subjects of predication in SRs, as shown in Section 3.3. Recall that depictive readings are allowed in Spanish with definite arguments (), but they are disallowed with bare nouns, since bare nouns lack D (see Laca , on this issue). It has been said that null arguments that are visible as subjects of predication must project D (). Hence, DNOs display D features (i.e., D must be projected). In sum, if DNOs are projected as nP (i.e., they lack D and Num), the properties of DNOs in SRs cannot be explained.
According to Weir (), who follows Raposo () on object drop in European Portuguese, DNOs in recipes are instances of a null D and a null NP. That is why null objects in recipes are interpreted as definite and specific. This analysis relies on a null D in the recipe lexicon. Note that the grammar of articles in recipes is different from that of ‘neutral’ English. Even when lacking articles, the DPs in (14) are interpreted as definite and specific, as DNOs in recipes.
According to Weir (), the null Ds in (14) share semantics with the indefinite determiner some in English, because it licenses null complements, like the null article in DNOs (e.g., I brought some .vegetables)). Hence, the null articles in (14) should also license null complements. A null D licenses the null NP in (15).
However, the approach faces some problems. First, it is unclear why a null D exists in the recipe variety of English, given that article drop is not commonly found in overt arguments in recipes. Why is article drop only available for DNOs in recipes? The answer to this question is not clear. Second, the availability of null Ds in Spanish is not empirically adequate. The absence of articles leads to indefinite, non-specific readings inSpanish (Laca , , ). Contrary to English bare nouns (see for an overview), there is no context in which bare nouns can be interpreted as definite in Spanish. Therefore, the indefinite interpretation of DPs without articles is not consistent with one of the conditions to license null objects in SRs: their obligatory definite, specific interpretation. Additionally, the Spanish versions of the examples in (14) are ungrammatical with the same definite, specific reading as that in (15), as shown in (16).
Furthermore, if null objects in recipes were indefinite DPs, they could not behave as anaphoric expressions, given that indefinite DPs cannot refer to antecedents. Instead, they introduce new referents (Laca , , ). This is not consistent with one of the conditions that license DNOs in SRs: their obligatory referentiality. In sum, although the proposal is suggestive for DNOs in English recipes, it should be ruled out for DNOs in SRs.
The basic grammatical properties of DNOs in SRs are presented in this section. DNOs in SRs must be affected (3.1), must occur within directive utterances (3.2), and are syntactically projected, as they can control and bind grammatical elements, license parasitic gaps, and be visible as subjects of predication (3.3). Interestingly, DNOs without linguistic antecedents are also allowed in SRs (3.4).
One condition to license DNOs in SRs is related to affectedness. Intuitively, a participant in an event is affected when it suffers (or experiences) a change of state (physical or mental) as a result of the event (; ; Beavers 2008). In this sense, DNOs in SRs must be affected, as shown in (17).
‘Crush it for a few seconds.’
‘Seed it and cut it up.’
‘We season, pour and reserve it.’
Crucially, DNOs in SRs usually are patients, because patients tend to be affected. The tendency of DNOs to behave as patients predicts that non-affected DNOs should not be allowed in recipes. Thus, sentences like the one in (18), in which the DNO is not affected (the DNO is a theme), are not allowed ().
In light of these facts, one constraint for the occurrence of DNOs in recipes is that DNOs must be affected. That is why they are typically interpreted as patient-like entities.
Another constraint on the occurrence of DNOs in SRs is that they must occur within directive utterances. Therefore, the availability of definite object drop in (6) is not only related to the recipe context (pace ), but also to certain grammatical conditions, given that definite object drop is not permitted in the examples in (19), also from recipe contexts. This is so because, in the absence of imperative values, the grammar of the ‘neutral’ Spanish register (i.e., the grammar that does not license DNOs) applies in (19). The same applies to the English cases, as observed in Weir ().
‘I boiled the chicken breasts for 15’. I removed, and drained them.’
‘I will add the water and boil it.’
‘I cleaned the mussels. I got them out of the shell.’
Recall that imperative mood is not only expressed by (canonical) imperative forms, but also by inflected plural forms and infinitives (; ; ). Thus, DNOs in SRs can occur with ‘canonical’ imperatives, as shown in (20a), but also with infinitives and inflected plural forms with directive values, as illustrated in (20b) and (20c) respectively. From a pragmatic point of view, contexts licensing DNOs in SRs are linked to directive utterances (i.e., commands).
‘Cook the chicken breasts for 15’. Remove, and drain them.’
‘Add the water and boil it.’
‘We clean the mussels. We get them out of the shell.’
Therefore, in view of the data, DNOs are not only licensed by the recipe context or the prominence of the antecedent (pace ). Additional grammatical reasons, specifically that DNOs must be selected by imperative verbs (more broadly, by expressions that permit directive interpretations), should be taken into account.
A crucial question is why imperatives (and, in general, directive expressions) can license the ellipsis sites in (20), but inflected verbs (without directive interpretations) cannot in (19). The tendency of imperatives to license ellipsis sites has been observed in the literature (see on subjects; on TP; for criticism). In fact, the imperative context is one of the few to permit null subjects in a non-null-subject language like English (; . From a pragmatic perspective, one of the reasons why imperatives favor the occurrence of null arguments is that imperatives are sensitive to the context of the speech act, since the person issuing the command coincides with the speaker, and the person carrying it out coincides with the addressee (i.e., the information about the participants is deictically obtained) (). From a formal perspective, imperatives only display second person forms, which makes the addressee identifiable (). Additionally, imperatives and infinitives have been said to lack T (; ). This implies the impossibility of assigning nominative case to the subject position (even though the subject is still conceptually present).
Regarding objects, imperatives (and, in general, directive utterances) can also trigger their omission. In principle, it would be desirable to extend the idea mentioned above to the case of DNOs in recipes (i.e., definite object drop in Spanish is only licensed when T is absent). However, objects are not governed by T in Spanish, and the example in (20c) rules out this possibility, since the verb is inflected. Therefore, the occurrence of DNOs is not related to the absence of T, but to the directive value of the utterance.
Additionally, note that if there is no concrete grammatical feature (related to imperative morphology) to explain these cases, but an interpretation that can be obtained otherwise (probably pragmatically), it is necessary to explain at what level this interpretation is obtained. From a cartographic approach, like that in Speas & Tenny (), it is an option to propose a functional projection for these cases (i.e., Addressee Phrase or Speaker Phrase). However, this proposal faces two problems. First, non-finite sentences and all sentences inflected in first person plural forms would be systematically ambiguous between two structures (i.e., imperative structure and non-imperative structure), but this is not the case. Second, entities that DNOs refer to do not identify as speaker nor as addressee. In fact, addressees are typically linked to (null) subjects, rather than to (null) objects. Thus, functional projection cannot explain the behavior of DNOs in recipes. The concrete properties of directive utterances that trigger the omission of the object are still not clear. For reasons of space, they are left for future research. The basic idea is that there is no concrete syntactic condition, but a pragmatic juncture that produces the effect described above. This pragmatic juncture is basically built upon the instructional register itself, which makes the discourse referents quite prominent by specifying the ingredients in a recipe list, and also by deictic means ().
As DNOs in SRs are interpreted, but not spelled-out, one problem to discuss is if they are syntactically projected. A null argument is projected if it can make grammatical relations with other constituents. The main ways to test this, mentioned in the literature, are control, binding, parasitic gaps and predication (; ).
First, DNOs in SRs can control the interpretation of null subjects, typically PRO (; ; ; but see and for criticism). The DNOs in (21) control the definite interpretation of each PRO. This is why DNOs and PRO are coindexed (). This control relation is only possible on the assumption that the DNO is projected in syntax.
‘Let it cool a little bit.’
‘We let it boil for 30’.’
Furthermore, DNOs introduce new referents that can be bound by anaphoric expressions. The null subject of se funda ‘melts’ refers to the DNO of revolver ‘to mix’ in (22). Thus, the null subject and the DNO are coreferential (). This anaphoric relation is only possible on the assumption that the DNO is projected in syntax.
[Referring to the butter] Revolver Øi bien para que Øi se funda.
‘Mix it well until it melts.’
Additionally, DNOs are crucial for parasitic gaps to ‘survive’. Parasitic gaps are empty categories whose occurrence depends on another gap filled with an empty category (; ; ; Postal , ; ). The possibility that there is a parasitic gap in (23a) is confirmed by the ungrammaticality of (23b), where the parasitic gap cannot be licensed by a previous gap (a clitic should occur). This shows that the DNO of reservamos ‘we reserve’ in (23a) must be syntactically projected, since the existence of the parasitic gap depends on the occurrence of the DNO.
Additionally, DNOs are visible as subjects of predication. The adjective templada ‘warm’ receives a depictive reading and shows feminine gender and singular number forms in (24). This shows that DNOs in SRs must be syntactically projected, since the adjective must agree with some syntactically projected element in the sentence (i.e., the DNO).
[Referring to the salad] Se recomienda servir Ø templada.
‘It is recommended to serve it warm.’
The above tests prove that DNOs are projected phrases that participate in grammatical relations like control, anaphora, structures with parasitic gaps, and predication.
It has been stated in the literature that DNOs in recipes “need not have a particular linguistic antecedent” (). Interestingly, examples of DNOs without linguistic antecedents can be found in SRs, as those of English recipes in Ruda (). For instance, the DNO of enfriar ‘to cool’ in (25a) is not interpreted as referring to the split antecedent ‘la sal, el vinagre y el aceite’ (; ), but as the product of those ingredients together (i.e., a set of mixed ingredients), which has been formed by passing the ingredients through a strainer. Thus, the DNO obtains its interpretation from a non-linguistic antecedent (i.e., a contextually created constituent). The same can be applied to (25b). Note that the new mixture of ingredients is implied when a causative verb occurs, transforming previous ingredients into new ones.
‘Add salt, vinegar and oil. Pass it through a strainer and leave it to cool.’
‘Remove the basil and the kaffir leaf, crush and strain it. Reserve it warm.’
A piece of evidence that the antecedent of the DNO of reservar ‘to reserve’ in (25b) is not split is related to predication. Crucially, the depictive tibio ‘warm’ in (25b) is not inflected for feminine gender. If the antecedent were split (i.e., ‘la albahaca y la hoja de kaffir’), the depictive would display feminine gender (e.g., tibia ‘warm’), as do its antecedents.
Additionally, the existence of ellipsis sites with non-linguistic antecedents rules out an analysis in terms of ellipsis, given that ellipsis must refer to some material mentioned in the discourse (; ; ).
This section summarizes the ideas presented so far. First, with regard to previous analyses, DNOs in SRs cannot be derived by wh-traces, since DNOs are implausible to occur within islands. They cannot be derived either by NP-traces, since the subject position is occupied by a projected subject (i.e., a null operator cannot occupy that position). They are not instances of nP either, because DNOs display gender, number and D features (i.e., DP must be projected). Finally, they cannot be derived by article drop, since the absence of articles triggers indefinite, non-specific readings. Therefore, the accounts for DNOs in English recipes cannot be extended to the cases of DNOs in SRs.
Second, the conditions to license DNOs are the following. i) DNOs must be interpreted as definite and specific. This interpretation can be obtained from an antecedent or by lexical or pragmatic means. ii) DNOs represent arguments that could occur as weak pronominals (i.e., clitics), with the same definite, specific reading. iii) DNOs must occur within directive utterances. iv) DNOs must be affected.
Third, regarding the syntactic properties of DNOs in SRs, they are syntactically projected as full-fledged DPs, because they can control and bind grammatical elements, parasitic gaps depend on them and they are visible as subjects of predication.
Since previous accounts for English DNOs in recipes cannot account for DNOs in SRs, the necessity of a novel proposal is motivated. This proposal should capture the above-mentioned conditions for DNOs to occur in SRs, along with their grammatical properties. In Section 5, it is argued that DNOs in SRs behave as null pronominals, in line with Farrell (), Kato () and Dimitriadis (, ).
This section discusses the idea that DNOs display pronominal behavior, and suggests that DNOs in SRs are better understood as non-spelled out pronouns (5.1). Recall that pro is discarded, since it is governed by T, typically in subject position (Chomsky , ; ). Finally, the crucial predictions of this analysis are presented (5.2).
Let us assume, in line with Ruda (2014100), that DNOs in recipes display pronominal behavior. Specifically, DNOs in SRs display the behavior of pronominal clitics (; Zwicky , ; ) due to three reasons.
First, the fact that DNOs and clitics alternate suggests that DNOs display pronominal behavior. Recall that this alternation is not only syntactic, but also semantic (i.e., DNOs and clitics receive the same interpretation in recipes). This shows that DNOs and clitics should share basic properties. Other cases of null objects in Spanish, like indefinite null objects, cannot alternate with clitics, at least not with the same interpretation (see for further insights). This shows that indefinite null objects do not display pronominal behavior. This is exactly the opposite case to that of DNOs in SRs. Additionally, DNOs and clitics also alternate in earlier stages of acquisition, as stated for French in Grüter (2006) and for (European) Portuguese in Costa & Lobo (). This shows that the alternation between these two uses is also present in children, therefore the two grammatical mechanisms must share some basic properties, differing only in their spell-out properties.
Second, DNOs in recipes must obtain their interpretation from an antecedent or by lexical or pragmatic means (when retrieving non-structurally coded information). From a formal perspective, this is explained by assuming that object pronouns (i.e., clitics) display a D head and an NP whose interpretation is obtained from an antecedent (). Concretely, D is the feature that shows referential ability (this is why bare nouns cannot refer to antecedents). Hence, clitics cannot denote entities by themselves. This is the exact behavior of DNOs in SRs.
Finally, clitics and DNOs are weak grammatical elements that display the behavior of topics. Weak grammatical elements (i.e., clitics and null arguments) mark high accessibility. This is shown in the Accessibility Hierarchy (), represented in (26), where the main types of DPs appear arranged along a scale that goes from the markers of high accessibility (more to the left side), to those of low accessibility (more to the right side) (see ; for other versions of the scale).
If the antecedent is accessible, null arguments or clitics are typically used in Spanish to refer to that antecedent, following (26). In ‘neutral’ registers of Spanish, if an antecedent is definite, it cannot be referred to with null objects (i.e., clitics are used instead). In SRs, null objects can be used for the same task, in alternation with clitics. Note that it makes sense to alternate between a null object and a clitic, since they are on the left side of the scale in (26). Hence, the alternation between DNOs and clitics is a consequence of the distribution of the elements in (26). Therefore, it seems reasonable to conclude that DNOs display pronominal behavior, in line with Ruda (). However, contrary to Ruda (), DNOs in SRs are projected as full-fledged DPs, given their syntactic representation.
The view according to which DNOs are pronominal entities is advantageous to capture the grammatical properties of DNOs. First, regarding their interpretation, null objects in SRs must bear definite, specific readings. Therefore, if they are considered to behave as non-spelled-out pronominals, whose interpretation is obligatorily definite and specific (; ), this basic property is naturally captured.
Second, if the DNOs in (27) are projected as null pronominals in object position (i.e., clitics), it is expected that gender and number of DNOs are not inherent features, but they are just reproduced from the antecedent. Each DNO in (27a) is represented as a feminine, plural pronominal, agreeing with the feminine plural antecedent. These features are also shown in the explicit clitic in córtalas ‘cut them’. The same can be applied to (27b-c). Thus, DNOs display gender and number features reproduced from an antecedent, in line with the behavior of clitics in Spanish, like las ‘them’ in (27a).
‘Boil the chicken breasts for 15’. Remove, drain, and cut them into fine slices.’
‘Remove the stem, deseed and chop it.’
‘We toast the slice of bread and die-cut it in a round shape.’
Third, the proposal captures that DNOs are visible as subjects of predication. Recall that depictive readings are only obtained with definite arguments (). Thus, the adjective tibia ‘warm’ receives depictive readings in (28a-b), but such a reading is not obtained with the indefinite DP pechugas ‘chicken breasts’ in (28c). Recall that, according to Landau (), null arguments that are visible as subjects of predication display D features. Therefore, DNOs display D features, like pronominals. Additionally, if DNOs would not display D, they could not refer to antecedents, since bare NPs (i.e., bare nouns), like the one in (28c), introduce new referents (Laca , , ).
Note also that the DNO in (29) can be replaced by a clitic, with the same depictive reading of templada ‘warm’. This is coherent with the idea that DNOs alternate with clitics in SRs, and that DNOs are interpreted as pronominal clitics (i.e., as definite, specific DPs).
[Referring to the salad] Se recomienda servir(la) templada.
‘It is recommended to serve it warm.’
Fourth, the proposal captures the fact that DNOs can lack linguistic antecedents, as shown in (30). As said above, DNOs in these cases obtain their interpretation not by a structural antecedent, but by other means (i.e., by contextually created constituents).
‘Add salt, vinegar and oil. Pass it through a strainer and let it cool.’
‘Remove the basil and the kaffir leaf, crush and strain it. Reserve it warm.’
If DNOs are projected as weak pronominals, the DNOs in (30) should also behave as clitics. However, if the DNO of colar ‘to strain’ in (30b) is projected as a clitic, it should refer to a prominent antecedent. The only prominent antecedents are la albahaca ‘the basil’ and la hoja de kaffir ‘the kaffir leaf’. In principle, if the DNO referred to them, it would be inexplicable how to strain basil and kaffir leaf, as these ingredients cannot be strained. As mentioned above (Section 3.4), it is necessary to assume that what gets strained is a new ingredient composed by the mixture of the previously crushed basil and kaffir leaf. The (apparent) problem is that, if this idea is assumed, the grammatical similarity between clitics and DNOs seems to be unworkable, since clitics refer to linguistic antecedents.
So far, clitics have been considered to behave as grammatical entities that refer to antecedents, from which they ‘reproduce’ gender and number features. However, the accusative clitic of third person can be derived from the Latin neuter form ILLUD. That is why there exists a neuter lo ‘it’ in Spanish. This neuter clitic can refer to complements or propositions in Spanish (Fernández Soriano , ). In (31a), lo ‘it’ refers to a proposition, and in (31b), it refers to all the antecedentes in the discourse. In the latter case, the proposition ponlo ‘put it’ is interpreted as ponlo todo ‘put it all’, that is, as a neuter clitic plus quantifier float, also common in English (; ; ; ; ). Recall that, even though lo ‘it’ refers to several ingredients in (31b), it agrees in singular, contrary to what one would expect. This is due to the fact that neuter clitics lack ‘marked’ number features (). Note that the antecedents of lo ‘it’ in (31) bear no specification for gender and number features, since they are sentences (i.e., phrases with no ‘marked’ gender and number features) and multiple antecedents (i.e., phrases whose gender is masculine and feminine alternatively). This is the reason why the clitics displays the neuter form lo ‘it’.
The basic behavior of neuter clitics in Spanish is also present in SRs. The reading obtained when using lo ‘it’ in (32a) is similar to that of (32b). This is the case because, when using a causative verb like triturar ‘to crush’, a new ingredient is naturally created. This is why the neuter clitic lo ‘it’ is used. This way, the information about the new ingredient created from crushing lentils and raisins can be obtained. Recall that this reading has been already labeled as the ‘collective’ reading of neuter pronouns (). Additionally, the meaning of the proposition in (32a) can also be obtained with the neuter quantifier todo ‘everything’, as shown in (32b), and even with a neuter clitic and a quantifier float at the same time, as shown in (32c). This suggests that the same meaning can be displayed by three grammatical elements, as shown in (32).
‘I have crushed the lentils and the raisins, and I have put it in the frying pan.’
papel absorbente, filmar Ø y guardar Ø.
‘Chop the wings and the cocochas, and reserve for the filling. Dry everything
on absorbent paper, film it and save it.’
‘I have crushed the lentils and the raisins, and I have put all in the frying pan.’
Note that the masculine plural clitic los ‘them’ can also be used in these cases, as shown in (33a), but the interpretation obtained is different from that of the neuter use of lo ‘it’ in (33b). In (33a), crushed ham and crushed rye are put separately into the frying pan, while in (33b), the mixture of both ingredients is put at the same time into the frying pan.
‘I have crushed the ham and the rye, and then I have put them in the frying
pan.’
‘I have crushed the ham and the rye, and then I have put it in the frying pan.
Additionally, analyzing these cases as neuter clitics accounts for the gender features of certain depictives. The adjective tibio ‘warm’ is predicated from the DNO of reservar ‘to reserve’ in (34a). If this DNO referred to la albahaca ‘the basil’ and la hoja de kaffir ‘the kaffir leaf’, the depictive would agree in feminine. However, it displays non-marked gender, which can be accounted for by representing the DNO as a neuter clitic, like that in (34b).
‘Remove the basil and the kaffir leaf, crush and strain it. Reserve it warm.’
‘Remove the basil and the kaffir leaf, crush them and reserve it warm.’
Further evidence comes from similar cases in English. Crucially, the pronoun it in English can function similarly to the pronoun lo ‘it’ in Spanish. The referent of it is not a structural entity (i.e., it is not syntactically projected) in (35), but rather an element created ‘post-semantically’ (see on it-anaphors; ).
Put [the yogurt]i, [cream]j, [sugar]k and [zest]z in a bowl; whip themi+j+k+zto soft
peaks. Spread it over the meringue.
In sum, the behavior of DNOs in SRs can be accounted for by assuming that they display the behavior of weak pronouns, in line with Grüter () for French null objects. Concretely, DNOs are projected as null pronominals, in line with previous works by Farrell (), Kato () and Dimitriadis (, ). If the antecedent displays masculine/feminine gender, the DNO is projected as a masculine/feminine clitic. If the DNO receives the reading of a new ingredient produced, the DNO is projected as a neuter clitic, its antecedent being selected by a causative verb. This is the case because, given that that the new ingredient is not grammatically represented, it cannot display gender and number features. Hence, these features cannot be reproduced by a clitic, which is inflected in non-marked forms (i.e., neuter and singular forms).
As opposed to ‘neutral’ registers of Spanish, definite object drop can be found under specific conditions in special contexts, such as recipes. These conditions are the following. i) Definite null objects must be interpreted as definite and specific. This interpretation can be obtained from an antecedent or by other means (i.e., lexical or pragmatic means). ii) Definite null objects represent arguments that could occur as weak pronouns (i.e., clitics), with the same definite, specific reading. iii) DNOs must occur within directive utterances. iv) Definite null objects must be affected.
Definite null objects are syntactically projected, given that they can control and bind grammatical elements, parasitic gaps depend on them and they are visible as subjects of predication. These tests show that definite null objects display grammatical features, specifically gender, number and D (i.e., they are projected as full-fledged DPs). In order to account for the grammatical properties and the pronominal behavior of definite null objects in Spanish recipes, it has been proposed in this paper that they are projected as null pronominals in object position. This analysis predicts a) the obligatory definite and specific readings of null objects in Spanish recipes, b) the alternation of null objects and clitics, since they are grammatical mechanisms used to refer to prominent antecedents, c) the fact that definite null objects are visible as subjects of predication, and d) why definite null objects can lack linguistic antecedents in recipes.
Nonetheless, there are some remaining issues worth addressing in future research. First, it is still unclear why the alternation is only possible in recipe contexts (and, more broadly, in instructional contexts). Second, although it has been proposed in this paper that definite null objects in Spanish recipes behave as pronominals, a formal analysis awaits an answer. Given the proposal, the analysis should fit the grammatical representation of clitics. Third, the tight relation between the occurrence of imperatives (and, in general, directive utterances) and the occurrence of definite null objects in Spanish recipes is not yet clear. Although there is no doubt that, given the pragmatic and formal properties of imperatives, the occurrence of certain null arguments, such as null subjects, is more frequently allowed with imperatives, it has not yet been clarified what the exact property of directives is that triggers the object omission in Spanish recipes. Finally, definite object drop can also be found in pragmatically controlled contexts, where the absence of the imperative makes it difficult to retrieve the reference of the null element () (e.g., [Pointing at an apple] ¡Dame Ø! ‘Give me!’). These uses parallel those of definite null objects in Spanish recipes in many senses. In principle, a unified approach is the preferred option from a theoretical perspective, but the information of the content of the object is obtained through different means in these cases (i.e., textually vs. deictically retrieved information). The natural connection between these two instances of definite object drop awaits an answer.
In sum, this paper has pointed out the existence of (European) Spanish definite object drop in recipes, contrary to the generalization that this kind of drop is not allowed in (European) Spanish, and has identified the grammatical properties of definite null objects in recipes. The analysis relies on a null pronominal in the ellipsis site. Future research might shed light on new data that strengthen the predictive capability of the proposal.
This research was funded by the Spanish Ministerio de Universidades, under the auspices of FPU20/04298, EST23/00059 and EST24/00098, by the financed project Variación gramatical del español: microparámetros en las interfaces de la sintaxis con los niveles morfológico-léxico y semántico-discursivo (PID2021-123617NB-C43, Spanish Agencia Estatal de Investigación), and by Departamento de Lengua Española y Teoría de la Literatura (Universidad Complutense de Madrid).
I am indebted to Victoria Escandell-Vidal, Cristina Sánchez López, Edita Gutiérrez Rodríguez, Esther Rinke, Alejo Alcaraz and Laura Stigliano for providing very helpful comments. Earlier versions of this paper were presented at Kolloquium für Examenskandidat*innen und Doktorand*innen (Goethe-Universität, November 6, 2023), at the panel La elipsis y la arquitectura del lenguaje: ¿Cuándo puede elidirse qué y por qué?, within LII Simposio de la SEL (CCHS-CSIC, January 25, 2024), at Seminario de Lingüística (Universidad Complutense de Madrid, February 5, 2024), at 27th OSUCHiLL (The Ohio State University, March 22, 2024) and at LIII Simposio de la SEL (Universidad de Salamanca, January 21, 2025). I thank the audience for the valuable suggestions. I am also grateful to the anonymous reviewers of Verba, whose comments have improved this work. All shortcomings are my own responsability.
1
Al Khalaf, Eman. 2019. Floating quantifiers are autonomous phrases: A movement analysis. Glossa 4, 1-23. https://doi.org/10.5334/gjgl.848.
2
Anderson, Mona. 2006. Affectedness. In Martin Everaert & Henk van Riemsdijk (eds.), The Blackwell Companion to Syntax, vol. 1, Oxford: Blackwell, 121-141. https://doi.org/10.1002/9780470996591.ch5.
6
Barbosa, Pilar. 2019. Pro as a Minimal nP: Toward a Unified Approach to Pro-Drop. Linguistic Inquiry 50, 487-526. https://doi.org/10.1162/ling_a_00312.
7
Beavers, John. 2011. On affectedness. Natural Language & Linguistic Theory 29, 335-370. https://doi.org/10.1007/s11049-011-9124-6.
8
9
Benmamoun, Elabbas. 1999. The Syntax of Quantifiers and Quantifier Float. Linguistic Inquiry 30, 621-642. https://doi.org/10.1162/002438999554237.
10
Bordelois, Ivonne. 1986. Parasitic gaps: Extensions of restructuring. In Ivonne Bordelois & Heles Contreras (eds.), Generative Studies in Spanish Syntax, Berlin: De Gruyter, 1-24. https://doi.org/10.1515/9783110859232-003.
11
Bošković, Željko. 2010. On NPs and clauses. In Günther Grewendorf & Thomas Ede Zimmermann (eds.), Discourse and grammar: From sentence types to lexical categories, Berlin: Walter de Gruyter, 179-242. https://doi.org/10.1515/9781614511601.179.
12
13
Bosque, Ignacio; Brucart, Josep M.ª 2019. Caribbean Spanish and theoretical syntax. An overview. In Ángel Gallego (ed.), The Syntactic Variation of Spanish Dialects, Oxford: Oxford University Press, 297-328. https://doi.org/10.1093/oso/9780190634797.003.0011.
15
16
Büring, David. 2005. Binding Theory, Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9780511802669.
17
18
19
Camacho, José. 2013. Null Subjects, Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9781139524407.
21
22
Cardinaletti, Anna. 2002. Against optional and null clitics. Right dislocation vs. marginalization. Studia Linguistica 56, 29-57. https://doi.org/10.1111/1467-9582.00086.
23
24
26
27
28
Clements, J. Clancy. 2006. Null Direct Objects in Spanish. In J. Clancy Clements & Jiyoung Yoon (eds.), Functional Approaches to Spanish Syntax, London: Palgrave Macmillan, 134-150. https://doi.org/10.1057/9780230522688_6.
29
Contreras, Heles. 1993. On null operator structures. Natural Language & Linguistic Theory 11, 1-30. https://doi.org/10.1007/BF00993019.
30
31
32
Cyrino, Sonia. 2016. The null object in Romania Nova. In Mary A. Kato & Francisco Ordóñez (eds.), The Morphosyntax of Portuguese and Spanish in Latin America, Oxford: Oxford Studies in Comparative Syntax, 177-203. https://doi.org/10.1093/acprof:oso/9780190465889.003.0008.
33
34
35
Culy, Christopher. 1996. Null objects in English recipes. Language Variation and Change 8, 91-124. https://doi.org/10.1017/S0954394500001083.
36
37
38
39
Farrell, Patrick. 1990. Null objects in Brazilian Portuguese. Natural Language & Linguistic Theory 8, 325-346. https://doi.org/10.1007/BF00135617.
40
41
42
Fernández Soriano, Olga. 2016. Clíticos. In Javier Gutiérrez-Rexach (ed.), Enciclopedia de Lingüística Hispánica, Abingdon, Oxon: Routledge, 423-436. https://doi.org/10.4324/9781315713441-38.
43
Fillmore, Charles J. 1986. Pragmatically Controlled Zero Anaphora. Proceedings of the Twelfth Annual Meeting of the Berkeley Linguistics Society, 95-107. https://doi.org/10.3765/bls.v12i0.1866.
44
45
46
Flores, Cristina; Rinke, Esther; Sopata, Aldona. 2020. Acquiring the distribution of null and overt direct objects in European Portuguese. Journal of Portuguese Linguistics 19, 1-20. https://doi.org/10.5334/jpl.239.
47
48
49
Gómez Seibane, Sara; Camus Bergareche, Bruno. 2015. Nuevos datos acerca de la omisión de objetos en el castellano del País Vasco. Círculo de Lingüística Aplicada a la Comunicación (CLAC) 61, 211-236. https://doi.org/10.5209/rev_CLAC.2015.v61.48473.
50
Groefsema, Marjolein. 1995. Understood arguments: A semantic/pragmatic approach. Lingua 96, 139-161. https://doi.org/10.1016/0024-3841(95)00002-H.
51
Grüter, Theres. 2009. A Unified Account of Object Clitics and Referential Null Objects in French. Syntax 12, 215-241. https://doi.org/10.1111/j.1467-9612.2009.00128.x.
52
Gundel, Jeanette K. 2010. Reference and Accessibility from a Givenness Hierarchy Perspective. International Review of Pragmatics 2, 148-168. https://doi.org/10.1163/187731010X528322.
53
Gundel, Jeanette K.; Hedberg, Nancy; Zacharski, Ron. 1993. Cognitive Status and the Form of Referring Expressions in Discourse. Language 69, 274-307. https://doi.org/10.2307/416535.
54
55
56
Haegeman, Liliane. 1987b. Register variation in English: Some theoretical observations. Journal of English Linguistics 20, 230-248. https://doi.org/10.1177/007542428702000207.
57
Haegeman, Liliane. 1990. Understood subjects in English diaries. Multilingua 9, 157-199. https://doi.org/10.1515/mult.1990.9.2.157.
58
Haegeman, Liliane. 1997. Register variation, truncation, and subject omission in English and in French. English Language and Linguistics 1, 233-270. https://doi.org/10.1017/S1360674300000526.
59
60
Haegeman, Liliane. 2013. The syntax of registers: Diary subject omission and the privilege of the root. Lingua 130, 88-110. https://doi.org/10.1016/j.lingua.2013.01.005.
61
Hornstein, Norbert. 1999. Movement and Control. Linguistic Inquiry 30, 69-96. https://doi.org/10.1162/002438999553968.
62
63
Jaeggli, Osvaldo; Safir, Kenneth J. 1989. The Null Subject Parameter and Parametric Theory. In Osvaldo Jaeggli & Kenneth J. Safir (eds.), The null subject parameter, Dordrecht: Kluwer Academic Publishers, 1-44. https://doi.org/10.1007/978-94-009-2540-3_..
64
Kato, Mary Aizawa. 1993. The distribution of pronouns and null elements in object position in Brazilian Portuguese. In William J. Ashby, Marianne Mithun, Giorgio Perissinotto & Eduardo Raposo (eds.), Linguistic Perspectives on the Romance Languages, Amsterdam: John Benjamins, 225-235. https://doi.org/10.1075/cilt.103.23kat.
65
66
67
Laca, Brenda. 2013. Spanish bare plurals and topicalization. In Johannes Kabatek & Albert Wall (eds.), New Perspectives on Bare Noun Phrases in Romance Languages and Beyond, Amsterdam: John Benjamins, 95-120. https://doi.org/10.1075/slcs.141.04lac.
68
Landau, Idan. 2010. The Explicit Syntax of Implicit Arguments. Linguistic Inquiry 41, 357-388. https://doi.org/10.1162/LING_a_00001.
69
70
71
72
73
74
Marchis, Mihaela & Alexiadou, Artemis (2013): The syntax of clitics revisited: Two types of clitics. Lingua 127, 1-13. https://doi.org/10.1016/j.lingua.2013.01.002.
75
76
Martins, Ana María. 1994. Enclisis, VP-deletion and the nature of Sigma. Probus6, 173-205. https://doi.org/10.1515/prbs.1994.6.2-3.173.
77
Massam, Diane. 1992. Null objects and non-thematic subjects. Journal of Linguistics 28, 115-137. https://doi.org/10.1017/S0022226700015012.
78
79
Masullo, Pascual José. 2017. La interfaz sintaxis-pragmática: Caída del objeto acusativo definido sin clítico en el español rioplatense. Saga Revista de Letras 7, 53-72. https://doi.org/10.35305/sa.vi7.45.
80
Merchant, Jason. 2001. The Syntax of Silence. Sluicing, Islands and the Theory of Ellipsis, Oxford: Oxford University Press. https://doi.org/10.1093/oso/9780199243730.001.0001.
81
Neeleman, Ad; Szendrői, Inger. 2007. Radical pro drop and the morphology of pronouns. Linguistic Inquiry 38, 671-714. https://doi.org/10.1162/ling.2007.38.4.671.
82
Okada, Sadayuki (1998). Reflexive pronouns with split antecedents. Journal of Pragmatics 30, 59-84. https://doi.org/10.1016/S0378-2166(97)00088-X.
83
84
Palacios, Azucena. 2000. El sistema pronominal del español Paraguayo: un caso de contacto de lenguas. In Julio Calvo Pérez (ed.), Teoría y práctica del contacto: el español de América en el candelero, Madrid: Iberoamericana/Frankfurt a. M.: Vervuert, 122-143. https://doi.org/10.31819/9783865278883-007.
85
86
Paul, Ileana; Massam, Diane. 2021. Licensing null arguments in recipes across languages. Journal of Linguistics 57, 815-839. https://doi.org/10.1017/S0022226721000293.
87
Platzack, Christer; Rosengren, Inger. 1997. On the subject of imperatives: A minimalist account of the imperative clause. Journal of Comparative Germanic Linguistics 1, 177-224. https://doi.org/10.1023/A:1009747522177.
88
89
91
92
93
Raposo, Eduardo. 1986. On the Null Object in European Portuguese. In Osvaldo A. Jaeggli & Carmen Silva-Corvalán (eds.), Studies in Romance Linguistics, Dordrecht: Foris, 373-390. https://doi.org/10.1515/9783110878516-024.
94
96
Ritter, Elizabeth. 1995. On the syntactic category of pronouns and agreement. Natural Language & Linguistic Theory 13, 405-443. https://doi.org/10.1007/BF00992737.
97
98
100
Ruda, Marta. 2014. Missing objects in special registers: The syntax of null objects in English. The Canadian Journal of Linguistics 59, 339-372. https://doi.org/10.1017/S0008413100000396.
102
Ruppenhofer, Josef; Michaelis, Laura A. 2010. A constructional account of genre-based argument omissions. Constructions and Frames 2, 158-184. https://doi.org/10.1075/cf.2.2.02rup.
103
Ruppenhofer, Josef; Michaelis, Laura A. 2013. Frames and the interpretation of omitted arguments in English. In Stacey K. Bourns & Lindsy L. Myers (eds.), Perspectives on Linguistic Structure and Context: Studies in honor of Knud Lambrecht, Amsterdam: John Benjamins. https://doi.org/10.1075/pbns.244.04rup.
105
106
107
Sánchez, Liliana. 1999. Null objects and Dº features in contact Spanish. In Jean-Marc Authier, Barbara A. Bullock & Lisa A. Reed (eds.), Formal Perspectives on Romance Linguistics, Amsterdam: John Benjamins, 227-242. https://doi.org/10.1075/cilt.185.17san.
108
110
111
Shimojo, Mitsuaki. 2019. Topicalization in Japanese cooking discourse. Open Linguistics 5, 511-531. https://doi.org/10.1515/opli-2019-0028.
112
Silva, María Florencia. 2023. Imperatives and their interaction with VP-ellipsis in Spanish. Isogloss 9, 1-16. https://doi.org/10.5565/rev/isogloss.262.
113
Speas, Peggy & Tenny, Carol L. 2003. Configurational properties of point of view roles. In Anna Maria Di Sciullo (ed.), Asymmetry in Grammar, vol. 1: Syntax and semantics, Amsterdam: John Benjamins, 315-344. https://doi.org/10.1075/la.57.15spe.
115
116
117
118
119
Weir, Andrew. 2012. Left-edge deletion in English and subject omission in diaries. English Languages and Linguistics 16, 105-129. https://doi.org/10.1017/S136067431100030X.
120
Weir, Andrew. 2017. Object drop and article drop in reduced written register. Linguistic Variation 17, 157-185. https://doi.org/10.1075/lv.14016.wei.
121
Wurmbrand, Susi. 2002. Semantic vs. syntactic control. In C. Jan-Wouter Zwart & Werner Abraham (eds.), Studies in comparative Germanic syntax. Proceedings from the 15th Workshop on Comparative Germanic Syntax, Amsterdam: John Benjamins, 93-127. https://doi.org/10.1075/la.53.07wur.
122
124
Zwicky, Arnold M. 1985. Clitics and Particles. Language 61, 283-305. https://doi.org/10.2307/414146.
[1] I use the symbol Ø to mark null elements, and square brackets to identify the antecedent of each null element. I mark with subscripts coreferential elements in each sentence. If no subscript appears, I am assuming that identity of sense anaphora is involved. Coreferentiality is typically found in definite object drop, while identity of sense anaphora is commonly found in indefinite object drop (see on anaphoric relations).
[2] Note that definite object drop is possible in some varieties of American Spanish (see on Quiteño Spanish; on Andean Spanish; on Paraguayan Spanish; on River Plate Spanish), and in Basque Spanish (see ; ). In most of the cases, Spanish is in contact with languages that license definite object drop, like Quechua, Basque and Guarani. Diatopic variation is not discussed here.
[3] An anonymous reviewer argues that it is not necessary to specify that definite DPs are specific, since all definite DPs are inherently specific. However, I think this distinction needs to be maintained, as has been pointed out in the literature that some (formally) definite DPs can receive indefinite and non-specific readings under certain circumstances (see Leonetti , for further information).
[4] Interestingly, this alternation is also found in languages licensing definite object drop in ‘neutral’ registers, like (European) Portuguese, which permits definite object drop with definite, inanimate antecedents (Raposo , ; ). The alternation in (European) Portuguese triggers no difference in meaning either. I leave aside here the possible correlation between these two phenomena.
[5] Note that definite null objects can also be found in sentences inflected in the first person singular (e.g., Cuezo las pechugas. Retiro Ø y escurro Ø ‘I cook the chicken breasts. I remove and drain them’). These cases can most frequently be found in recorded recipes (e.g., recipes on YouTube or Instagram). Even though recorded recipes are not the object of study of this paper, it would be very telling to analyze them in future research.
[6] As an anonymous reviewer correctly points out, definite null objects have been analyzed in terms of null clitics in the literature (see and on Brazilian Portuguese; Dimitriadis , on Modern Greek; on French). However, I avoid the term null clitic, and use the term null pronominal. According to Cardinaletti () and Leonetti (), the notion of null clitic is not reasonable, because clitics display phonological properties by definition. As this property is not compatible with the idea that clitics could be non-spelled-out entities, these authors prefer the term null pronominal instead.
Conversely, Dimitriadis (, ) claims that clitics are a recognizable syntactic class of pronominals, so the term null clitic is used to refer only to the syntactic properties of the class “without reference to its phonological properties” (). Apart from terminological issues, the crucial point is that definite null objects in Spanish recipes behave as (null) pronominals, not as elided arguments (see on Brazilian Portuguese indefinite object drop as nominal ellipsis), and are categorized as DPs.
Additionally, I fully agree with the ideas presented in and that definite null objects display pronominal behavior, but I do not delve into the issue of whether definite null objects in Spanish recipes should be analyzed as pro with inherent third person features, as argued for by these authors for Brazilian Portuguese. My concerns are whether definite null objects in Spanish recipes display pronominal behavior, what their features are and to which grammatical category they belong.
The examples given in this paper are based on of naturally occurring data. The web pages were accessed in 2023.
Available at: https://www.hogarmania.com/cocina/
(www.hogarmania.com/cocina/recetas/ensaladas-verduras/ensalada-cesar-receta-fama-mundial.html)
Pon agua a calentar en un cazo, agrega el huevo y cuécelo durante 10 minutos desde el momento en que el agua empiece a hervir. Retira Ø, refresca Ø, pela Ø, separa la yema de la clara y reserva Ø.
Salpimienta las pechugas de pollo, átalas (por separado) con lid de cocina (de manera que queden recogidas) y ponlas a cocer en una cazuela con agua sazonada con una pizca de sal. Añade también unas ramas de perejil. Cuece las pechugas durante 15 minutos. Retira Ø, escurre Ø, córtalas en filetes finos y reserva Ø.
Lava Ø, seca Ø, trocea las hojas de las lechugas y ponlas en una fuente grande. Retira la corteza del queso, córtalo en dados y resérvalo. Corta el pan en daditos y fríelos en una sartén con abundante aceite. Escúrrelos sobre un plato cubierto con papel absorbente de cocina.
Para hacer el aliño, pica el diente de ajo y ponlo en el mortero. Pica las anchoas y agrégalas. Añade también la yema del huevo y maja todo bien. Incorpora la mostaza, el yogur, el zumo de limón, la salsa inglesa y un chorrito de aceite. Mezcla Ø bien y reserva Ø.
Monta las ensaladas colocando en la base de los platos las hojas variadas de lechugas y sazónalas. Reparte encima los filetes de pollo, los dados de queso y los costrones de pan frito. Aderézalas con el aliño y sirve Ø.
(https://www.hogarmania.com/cocina/recetas/sopas-cremas/gazpacho-andaluz-841.html)
Escalda y pela los tomates e introduce en el vaso de la batidora, añade el pepino pelado y troceado, el pimiento, el ajo y el pan remojado en agua.
Tritura Ø durante unos segundos y añade la sal, el vinagre y el aceite probando para rectificar si fuera preciso.
Pasa Ø por chino a una sopera y deja Ø enfriar durante una hora.
Sirve el gazpacho andaluz de Karlos Arguiñano en tazas individuales acompañados de una guarnición de dados de pan, de pepino, de cebolla, de tomate y de huevo duro.
(https://www.hogarmania.com/cocina/recetas/postres/mini-donuts-de-chocolate.html)
Pon la harina en un bol grande y desmenuza la levadura encima. Añade la sal, el agua, el azúcar, el huevo y la yema. Ralla encima un poco de cáscara de limón y mezcla los ingredientes con una lengua hasta que desaparezca la harina. Incorpora la mantequilla, mezcla de nuevo hasta que quede integrada y deja que repose durante 30 minutos aproximadamente (autólisis).
Espolvorea la encimera con un poco de harina, coloca la masa encima y amásala durante 5 minutos aprox. Ponla de nuevo en el bol, tápala con film de cocina y deja que fermente a temperatura ambiente hasta que doble su volumen (2 a 3 horas).
Saca la masa del bol, divídela en 12 porciones, redondéalas y deja que reposen durante 10 minutos aproximadamente.
Aplasta cada bolita, introduce el dedo índice en el centro y forma los dónuts. Colócalos sobre una bandeja de horno cubierta con papel de horno, tápalos con paño limpio y deja que fermenten durante 30-40 minutos.
Calienta a 150º-160º abundante aceite en una sartén. Introduce 6 donuts y fríelos hasta que el aceite deje de burbujear. Dales la vuelta y fríelos por el otro lado. Retira Ø y escúrrelos sobre un plato cubierto con papel absorbente. Repite el proceso con los otros 6 dónuts. Espera a que estén bien fríos.
Funde el chocolate negro al baño María o en el microondas.
Cubre una bandeja de horno con film de cocina. Coloca encima una rejilla de horno y extiende los dónuts. Nápalos con el chocolate y adorna la mitad con unas pepitas de chocolate blanco. Sirve Ø y adorna Ø con unas hojas de menta.
(www.hogarmania.com/cocina/recetas/postres/crepes-caseros-en-thermomix.html)
Añade todos los ingredientes en el vaso de la Thermomix y programa 20 segundos a velocidad 4.
Una vez lista la masa, ya puedes empezar a hacer tus crepes caseros de Thermomix.
Engrasa una sartén grande (o una sartén para crepes) con aceite o mantequilla y calienta Ø en el fuego. Una vez coja temperatura, coloca una porción de crep con ayuda de un cucharón y mueve la sartén para que se esparza la masa por todos los lados.
Deja que se cocine unos minutos y cuando empiece a tener un tono dorado y a despegarse los laterales, dale la vuelta y deja que se cocine por el otro lado.
Una vez listo, sácalo de la sartén y sirve los crepes de Thermomix en un plato. Rellena Ø con los ingredientes que más te guste ¡y a devorar!
(https://www.hogarmania.com/cocina/recetas/sopas-cremas/vichyssoise-15196.html)
Para el caldo de pollo, pon la zanahoria, la parte verde de los puerros, unas ramas de perejil, el muslo y la carcasa de pollo en una cazuela. Cubre Ø con agua, pon Ø a punto de sal y cuece todo durante 20-25 minutos. Cuela Ø y reserva Ø.
Limpia la cebolleta y los puerros. Pica todo y pon Ø a pochar en una cazuela con la mantequilla y un chorrito de aceite. Añade las patatas peladas y cascadas y rehoga Ø a fuego suave. Sazona. Es importante que no cojan color. Vierte el caldo de ave y deja Ø cocer unos 12-15 minutos.
Tritura la sopa con la batidora eléctrica y cuela Ø.
Añade la nata líquida mientras mezclas Ø bien con una cuchara. Deja Ø atemperar y mete Ø en el frigorífico para que se enfríe bien.
Sirve la vichyssoise en un plato hondo o tazón con un poco de cebollino picado por encima ¡y disfruta de una cremita fría espectacular!
Available at: https://www.cordonbleu.edu/home/es
(https://www.cordonbleu.edu/news/madrid-receta-chipirones-rellenos-ajos-confitados-coulis-pimiento/es)
Ajos confitados:
Envolver los ajos en papel aluminio (papillote) con el aceite, el agua, el laurel, el tomillo, el romero y la sazón. Hornear a 120ºC hasta que se hagan (aproximadamente media hora). Una vez cocidos, apartar los ajos y reservar el aceite del confitado para freír los chipirones.
Chipirones:
Eviscerar los chipirones, quitar y desechar la pluma. Separar la cabeza de los tentáculos, aprovechar la cococha, despegar las alitas del cuerpo y quitar su membrana negra, además de la del cuerpo. Dar la vuelta al tubo (cuerpo), limpiar Ø y secar Ø. Picar las alitas y las cocochas finamente, y reservar Ø para el relleno. Reservar los tentáculos para freír. Secar todo sobre papel absorbente, filmar Ø y guardar Ø en la nevera.
Coulis de pimiento:
Tostar las semillas de comino y cilantro, y machacar Ø. Lavar el pimiento y el tomate, quitar el pedúnculo, despepitar Ø y trocear Ø. Unir todos los ingredientes en un cazo, cocer Ø media hora a fuego muy suave. Quitar la albahaca y la hoja de kaffir, triturar Ø y colar Ø. Reservar Ø tibio.
Relleno y cocción de los chipirones:
Mezclar el picado de las alitas y las cocochas, previamente salteados, con los ajos confitados. Rellenar el cuerpo de los chipirones. Cerrar Ø con un palillo de madera. Sazonar Ø y dorar Ø a la plancha con un poco del aceite de ajo, y reservar Ø. Aparte, salpimentar los tentáculos, rebozar Ø con la sémola y freír Ø a 170ºC con el aceite de oliva de ajo durante algunos segundos. Escurrir Ø.
Presentación:
Disponer una capa fina de coulis en el plato. Colocar encima un chipirón cortado en dos, un tentáculo frito, algunas hojas de perejil fritas y un hilo de aceite de perejil.
(https://www.cordonbleu.edu/news/receta-frittata-verduras/es)
Precalentar el horno a 170°C.
Rallar el calabacín, el queso parmesano y el pecorino. Picar la chalota, el puerro y los pimientos verde y rojo.
Picar finamente el perejil.
Calentar a fuego medio la mitad del aceite de oliva en una sartén apta para horno y agregar la chalota, el puerro y los pimientos. Cocinar Ø durante 4 minutos y añadir el calabacín, y continuar cocinando Ø 5 minutos más. Sazonar Ø.
Reservar Ø en un bol para que se enfríe.
En otro bol batir los huevos junto con la leche, el perejil y el queso parmesano. Añadir los vegetales y mezclar todo.
Calentar el resto del aceite en la sartén a fuego medio y verter la mezcla de huevos y vegetales, distribuyendo las verduras de manera uniforme.
Espolvorear el pecorino rallado por encima.
Cocinar la frittata a fuego medio/bajo durante 4 minutos sin mezclar hasta que los bordes estén firmes pero el centro aún esté suave.
Colocar la sartén en el horno precalentado y cocinar Ø alrededor de 8 minutos hasta que se haga por completo.
Retirar Ø, enfriar Ø y servir Ø.
(https://www.cordonbleu.edu/news/receta-dia-chef-cous-cous-ensalada-garbanzo/es)
Cous cous y ensalada de garbanzos
Calentar el aceite de oliva en un cazo a fuego medio, y añadir el apio, la chalota y el ajo. Saltear Ø hasta que se ablanden (unos 5 minutos). Añadir el agua y hervir Ø.
Retirar del fuego, añadir el cous cous, tapar Ø y dejar Ø hasta que el cous cous se hinche (10 minutos). Remover Ø con un tenedor para evitar que se peguen los granos.
Poner el cous cous en un bol, añadir los garbanzos, la ralladura de limón en conserva, el perejil y las almendras.
Añadir el vinagre, y los aceites de oliva y canola en un bol, y emulsionar Ø para crear una salsa vinagreta o aliño de ensalada. Sazonar Ø con sal, pimienta negra y miel.
Zanahorias Heirloom asadas y aliño de yogur
Precalentar el horno a 160ºC. Sazonar las zanahorias y echar una cucharada de aceite de oliva. Disponer Ø sobre una bandeja de horno y asar Ø hasta que las zanahorias estén tiernas y un poco doradas en color. Retirar Ø del horno.
Poner el yogur, el zumo y la ralladura de lima en un bol y mezclar Ø.
Para servir
Extender un poco del yogur en la base del plato de presentación y luego cuidadosamente, repartir el cous cous y la ensalada de garbanzos encima y, justo encima, las zanahorias heirloom. Añadir una cucharada más de aliño de yogur y espolvorear Ø con chaat masala. Para terminar, disponer las hojas de menta, y salpimentar Ø.
Se recomienda servir Ø templada o a temperatura ambiente.
(https://www.cordonbleu.edu/news/receta-tarta-fraisier-fresas/es)
Genovés:
Calentar los huevos con la yema y el azúcar al baño María a 50ºC - 55ºC.
Montar los huevos con la yema y el azúcar en la batidora hasta que se enfríe completamente.
Incorporar la mantequilla fundida a 45ºC-50ºC y al final la harina tamizada.
Hornear Ø a 180°C durante 25 minutos aproximadamente
Jarabe:
Hervir el agua con el azúcar.
Retirar Ødel fuego y añadir el licor de fresas.
Crema mousseline:
Comenzar elaborando la crema pastelera. En una cacerola pequeña, calentar la leche junto con las semillas de vainilla. Reservar Ø en caliente.
Mezclar la maicena con el azúcar y añadir las yemas de huevo. Batir Ø bien e, inmediatamente, blanquear la mezcla con las varillas.
Añadir la leche mezclada con la vainilla, todavía caliente, sobre la mezcla de yemas, azúcar y maicena, y mezclar Ø bien.
Llevar Ø al fuego de nuevo hasta que comience a hervir, sin dejar de revolver Ø para evitar que se pegue. Mantener Ø en el fuego bajo en hervor durante unos 8 segundos.
Quitar Ø del fuego y añadir la mitad de la mantequilla. Revolver Ø bien para que se funda.
Enfriar Ø dentro de una bandeja. Cubrir Ø con papel film.
Batir la mantequilla restante hasta que adquiera textura pomada. Añadir la crema pastelera, previamente alisada, a la mantequilla en varias veces.
Batir Ø hasta obtener una crema perfectamente lisa y aireada.
(https://ww.cordonbleu.edu/news/receta-macaron-cafe-chocolate/es)
Desarrollo de la masa de macaron:
Desarrollo del relleno:
Dejar Ø cristalizar en la nevera hasta que adquiera una textura similar a la de una crema de cacao y avellana.
Available at: https://www.rtve.es/television/masterchef/2013/recetas/
(https://www.rtve.es/television/20220622/receta-sopa-coco-claudia-masterchef/2384138.shtml)
(https://www.rtve.es/television/20210510/receta-cazuela-garbanzos-arnau/2089390.shtml)
(https://www.rtve.es/television/20230430/receta-bocadillo-berenjena-alex-masterchef/2441431.shtml)
