Sluicing as a Syntactic Phenomenon

Sluicing represents that specific type of ellipsis, which may occur either in direct or indirect interrogative clauses. Wh-phrase is the main indicator of the ellipsis. Other features are usually understood from the clause. Only recently sluicing has gained wide interest among linguists worldwide, who have started studying it more carefully to get a better idea of the ellipsis. A particular interest arises in the works of Chung, Sandra, John Ross, James McCloskey, and William Ladusaw, who were among the first to denote sluicing in that frame as it is perceived nowadays (Sandra et al 1995). Practically all languages have sluicing, which is described at its best with the following instances. The “sluiced” phrases are outlined by means of subscripts and italicized font, whereas the antecedent to this type of ellipsis is marked by bold fort:

• Tony wants to drink some water, but he doesn’t have it he wants to drink.
• Lisa doesn’t eat tomatoes, but she cannot explain why she doesn’t eat tomatoes.
• Someone has drunk all juice. Sadly, I don’t have a clue who did it has drunk all juice.

These clauses give a good understanding of how sluicing works in indirect questions. However, it does not mean that this type of ellipsis is not a common indirect question. The following examples certify a wide usage of direct sluicing, which may occur in the majority of languages.

The girl is arriving in the city today. – Who is arriving in the city today?

The boy put a letter on the table. – What did he put on the table?

The instances of such sluicing have the sluiced phrases going after their antecedents. However, such type of sluicing can also be found after the antecedent as in the following instances.

Despite the fact that I do not have a clue why the movie got such bad reviews, the movie got such bad reviews.

When and in which way someone is supposed to do that job is difficult to understand, but someone is supposed to do that job.

The potential change this ellipsis can make in the statement mostly depends on the language, where it is applied. For example, in some of them sluicing leaves behind not only the wh-material but other words too (a so-called multiple remnant sluicing as Osborne calls it) (Osborne et al 2013). An interesting fact is that the author notes an extreme rarity of multiple sluicing in such international language as English is (stating that this may have been the main reason why this syntactic phenomenon was not studied before), however, it is still possible to form a sluiced structure in the following-type constructions:

Does somebody want to drink something? I would like to know who wants to drink what.

Something is creating somebody lots of troubles, unfortunately, it’s not possible to comprehend what is creating who lots of troubles.

This kind of statement is considered to be grammatically correct in such languages as German, Turkish, Russian, Japanese, and some others. However, in the English language, their adequacy appears to be quite controversial. Multiple sluicing represents a latent issue for syntax, as the subcontext seems to create a non-constituent.

John Ross (1969) is considered to be the father of sluicing theory as he was the first who conducted an in-depth analysis of the phenomenon. He suggested that sluicing occurs with a wh-phrase, which is followed by the removal of its sister constituent. Together with Merchant, Ross analyzed standard sluicing which presupposed the interrogative clause to be reduced only to the wh-phrase (Paul & Potsdam n. d.).

Several years after Ross’s theory, the study has been promoted more profoundly by Merchant, who presented the most inclusive discourse on sluicing as a type of the ellipsis (Merchant 2001). The analysis of fronting together with removal has been questioned many times when the examples of multiple sluicing are provided, nonetheless, because it is quite challenging how both wh-phrases may be moved or followed in such an unusual way that may enable the remaining part to be removed afterward.

Another investigation of sluicing, as well as multiple sluicing, accepts the catena to be a primary unit of syntax, as Osborne together with Putnam and Gross states (Osborne 2012). The catena is a part of syntactic examination which is directly linked to dependency grammars. It can be most aptly defined by Gross who says that catena is a word or phrase that is permanent regarding dominance (Gross 2011). The subcontext of all examples of sluicing, including multiple sluicing, are mostly catenae. To stress this point more, the author creates dependency trees of the following instances of multiple sluicing which can well indicate the point:

As in the previous examples, it becomes clear that both the sluiced phrase and the antecedent meet the criteria of catenae. The second instance proves the point from a little bit different angle: The antecedent to the sluiced phrase as well as the sluiced phrase on its own meets the criteria of catenae again.

However, except for the standard view on sluicing, there are many exceptions, which do not fit into the frame of already existing sluicing norms. For example, a Western Austronesian language (used by the population of Madagascar), Malagasy, which may become a good representative of all wh-in-situ languages, still it does not have the same sluicing structures as other wh-in-situ languages possess (Chinese, Japanese and Javanese (Adams 2003) in particular) (Keenan 1976). This means that despite Ross and Merchant were trying to create a universal explanation for sluicing, the norm cannot be adopted unequivocally due to the differences endemic to each language instilled with the cultural and historical background.

Paul & Potsdam state that sluicing cannot be even called a syntactic construction by itself. All languages with wh-in-situ cannot prove a different side to the movement plus removal understanding of sluicing in case their sluicing-like constructions may be created in some other ways (Keenan 1976). Thus, a statement in English that seems to contain a lot of sluicing phrases shall not have any English-like derivation, whereas Japanese and Chinese (Wei 2004), on the opposite side, do not even contain the movement which may rely on the ellipsis and feed the sluicing derivation.

Whatever the discordances among the linguists on the subject of sluicing, its importance is undeniable as it also attracts attention to two other phenomena that are closely related to sluicing, namely ellipsis, and syntactic dependencies. Sluicing is a comparatively modern field of research for many scientists who seek to study even the tiny details of the languages around the whole world. As all of these languages should be analyzed only in the context of their cultural and historical backgrounds, the results of an analysis can rarely coincide or be unanimous for all languages. On the contrary, the investigation shows that the more one analyzes a certain language, the more differences he finds, like this literature review proved with English so opposite to the Chinese and Javanese languages when the talk goes about sluicing norms and ways.

The special feature of the English language also lies in its combination of both wh-in-situ (Japanese as the most prototype-like language of this kind) and wh-movement with overt characteristics (especially in multiple wh-questions when a part f the sentence remains in-situ and another part undergoes overt movement of wh-phrase) which are so endemic for the Slavic languages (for example, the Romanian language has numerous overt wh-characteristics due to the constant influence from such Slavic languages as Polish and Ukrainian). After comparing all the sluicing function in these three different languages, Alexandra Teodorescu and Frederick Hoyt state that due to the fact that the characteristics of sluicing phrases are presumed to derive from those wh-phrases that are believed to be common for a language under discussion, the patterns of sluicing for Romanian, Japanese, and English are far from being the same despite a wide range of similarities that exist between the three of them (particularly, the ellipsis of a CP-part in underlying cleft structures as well as ellipsis of an IP-part which occurs after overt wh-phrase). Japanese is a bright representative of the first type whereas English indicates the features of the second (Hoyt & Teodorescu 2012).