Article body

Introduction

This paper focuses on translational shifts with respect to the demonstrative determiner in French and Dutch in parallel corpora. So far, three similar studies have been conducted for other language pairs. In her corpus study, Jonasson (1998) found that in one third of the cases, French demonstrative determiners were translated by definite articles in Swedish, and that a French demonstrative with a recognitial function was generally translated by a definite article in Swedish. Whittaker (2004) found that noun phrases featuring a demonstrative were considerably more frequent in the French target texts than in the Norwegian source texts. Whittaker argues that this finding may be explained by the explicitation hypothesis, a well-known translation mechanism (Baker 1993; Blum-Kulka 1986; Chesterman 2001; Klaudy 1993; Pym 2005; Toury 2001) defined as follows by Blum-Kulka (1986):

The process of interpretation performed by the translator on the source text might lead to a TL text which is more redundant than the SL text. This redundancy can be expressed by a rise in the level of cohesive explicitness in the TL text. This argument may be stated as “explicitation hypothesis,” which postulates an observed cohesive explicitness from SL to TL texts regardless of the increase traceable to differences between the two linguistic and textual systems involved. It follows that explicitation is viewed here as inherent in the process of translation

Blum-Kulka 1986: 19

Finally, the results presented by Goethals (2007), whose corpus contains more Dutch demonstratives translated by Spanish definite articles than vice versa, suggest that the Dutch demonstrative pronoun is more grammaticalized than its Spanish equivalent and that this divergence cannot be attributed to translation mechanisms or translator preference alone. The present corpus study focuses on the translation of demonstrative determiners in French and Dutch parallel corpora.

Table 1 summarizes the different forms of the demonstrative determiner in both languages under consideration, while table 2 presents the forms of the definite article:

Table 1

Forms of the demonstrative determiner in French and Dutch

Forms of the demonstrative determiner in French and Dutch

-> See the list of tables

Table 2

Forms of the definite article in French and Dutch

Forms of the definite article in French and Dutch

-> See the list of tables

The present study aims to identify the types of translation shifts that occur systematically and to explore the underlying mechanisms and semantic effects of this demonstrative shifting. In this respect, complete omissions, as in example (1),[1] would seem to result from translator preference as an inherent element of the translation process. In example (2), on the other hand, demonstrative shifting seems to be driven by paradigmatic implicitation of the head noun in the target text (from Doubistes to team), whereas in example (3), it seems impossible to attribute the shift from a demonstrative determiner in the French source text to a definite article in the Dutch target text to mechanisms inherent in the translation process. Possible linguistic, structural differences between the French and Dutch demonstrative determiner systems will therefore be studied in greater detail (see also Vanderbauwhede to appear).

The paper is organized as follows: first, the corpus used is presented. Second, the basic findings and the main research question are summarized. The next section of the paper attempts to account for the process of demonstrative shifting at three levels, namely translator preference related to translation universals at the level of the noun phrase, specific manifestations of translation universals within the noun phrase, and structural differences between the French and Dutch demonstrative determiner systems. The paper concludes with a detailed overview of demonstrative shifting in the corpus.

1. Corpus

For the present analysis, the Dutch Parallel Corpus, a 10-million-word parallel corpus comprising texts in Dutch, English and French with Dutch as a pivotal language (Paulussen, Macken et al. 2006), was used. The DPC project was coordinated by the University of Leuven Kortrijk Campus and University College Ghent, and was funded by the Dutch Language Union (Nederlandse Taalunie) within the framework of the STEVIN programme (Essential Speech and Language Technology Resources), which promotes research in Dutch language and speech technology.

The Dutch Parallel Corpus is a well-balanced corpus designed to include the widest possible range of written texts. Text types range from literary prose to non-fiction material, such as essays or newspaper, business, technical and policy texts. Quality control mechanisms include manual verification, spot checks and automatic control procedures. Another advantage of this corpus is its sentence alignment (Trushkina, Macken et al. 2008).

For this analysis, we use a well-balanced sub-corpus of the DPC, including 50,000 French words and 50,000 Dutch words in the source text and their translation into Dutch and French respectively. The entire sub-corpus of 200,000 words consists of 40% essayistic texts, 40% newspaper texts (news reports, comments) and 20% informative texts (external communication). These different text types ensure a varied sub-corpus. However, in this paper, we will not focus on possible divergences between the three text types.

Table 3

The sub-corpus based on the Dutch Parallel Corpus

The sub-corpus based on the Dutch Parallel Corpus

-> See the list of tables

2. Basic findings and research questions

Table 4 and Table 5 give an overview of the translation of the French and Dutch demonstrative determiner in the sub-corpus:

Table 4

The translation of the demonstrative determiner in the DPC sub-corpus (French SL – Dutch TL)

The translation of the demonstrative determiner in the DPC sub-corpus (French SL – Dutch TL)

-> See the list of tables

Table 5

The translation of the demonstrative determiner in the DPC sub-corpus (Dutch SL – French TL)

The translation of the demonstrative determiner in the DPC sub-corpus (Dutch SL – French TL)

-> See the list of tables

First, overall, we find that 48.36% of the demonstrative determiners are translated by a demonstrative in the target text, and that this is the case in both directions (French – Dutch and Dutch – French). In 20.02% of the cases, a demonstrative determiner is translated by a definite article, or vice versa, and 31.58% of the demonstratives are translated by another word category (e.g., by an indefinite determiner, an adverb, a personal pronoun) or vice versa (i.e., another structure is translated by a demonstrative determiner). The divergences suggested by these quantitative results confirm the findings of a prior analysis (Vanderbauwhede 2007) based on a sub-corpus of another parallel corpus, i.e., the Namur Corpus (Paulussen 1999). In this sub-corpus (10% of the entire Namur Corpus), no less than 33% of the French and Dutch demonstratives were translated by a definite article or vice versa.

Second, there are several differences between French and Dutch. In the French SL corpus (443 demonstratives), the demonstrative determiner is more frequent than in the Dutch SL corpus (399 demonstratives). This is also the case for the French TL corpus (452 demonstratives) with respect to the Dutch TL corpus (330 demonstratives). Moreover, regardless of the direction between SL and TL, French demonstratives are much more frequently translated by the Dutch definite article than vice versa (i.e., 2.48/1), while French demonstratives are more commonly translated by another structure than vice versa (i.e., 1,54/1):

Table 6

Relative frequency of the demonstrative – definite article and demonstrative – other structure pairs in the DPC sub-corpus

Relative frequency of the demonstrative – definite article and demonstrative – other structure pairs in the DPC sub-corpus

-> See the list of tables

As the demonstrative determiner is underused in the Dutch TL corpus (330 demonstratives) compared with the French SL corpus (443 demonstratives) and overused in the French TL corpus (452 demonstratives) compared with the Dutch SL corpus (399 demonstratives), we can conclude from a methodological perspective that not only the French and the Dutch SL corpus, but also the French and the Dutch TL corpus can be considered as validated comparable corpora.

These important quantitative differences raise the following question: are the translational shifts of the French and the Dutch demonstrative determiner in the corpus due to translator preference related to translation universals at the level of the noun phrase (3.1.), to specific manifestations of translation universals within the noun phrase (3.2.), or to major structural differences between the French and Dutch demonstrative determiner systems (3.3.)? These possible explanations will now be examined in greater detail at the three different levels.

3. Explanations at three different levels

All divergent translations with respect to the demonstrative were collected and classified into three categories, namely divergences at the level of the noun phrase, at the level of the head noun and its modifiers, and at the level of the determiner. In what follows, it will be shown that this subdivision at three levels allows us to account for demonstrative shifting in the corpus.

3.1. Translator preference related to translation universals at the level of the noun phrase

Table 7, Table 8 and Table 9 show that changes at the level of the noun phrase are mainly due to either the complete omission or addition of the noun phrase, or its reformulation (i.e., reformulation of the whole utterance, adverbs, demonstrative pronouns and personal pronouns):

Table 7

Overview of changes at the level of the noun phrase (French SL – Dutch TL)

Overview of changes at the level of the noun phrase (French SL – Dutch TL)

-> See the list of tables

Table 8

Overview of changes at the level of the entire noun phrase (Dutch SL – French TL)

Overview of changes at the level of the entire noun phrase (Dutch SL – French TL)

-> See the list of tables

Table 9

Overview of differences at the level of the entire noun phrase in the DPC sub-corpus

Overview of differences at the level of the entire noun phrase in the DPC sub-corpus

-> See the list of tables

In the corpus, omissions or reformulations of the noun phrase with a demonstrative determiner in the target text are almost exclusively due to translator preference related to the broader context of translation universals, as a NP-to-NP translation would also have been possible. This is illustrated in the examples below.

Example (4) is an example of complete omission (32 cases): à ce moment-là has not been translated in the target text, but the translator could have translated this accessory phrase by the prepositional phrase op dat moment or by the adverb toen in Dutch. The temporal adverb toen does not contain any demonstrative determiner, but expresses the same type of anteriority as op dat moment and à ce moment-là.

Conversely, example (5) is an example of addition or explicitation (25 cases), a well-known phenomenon in translation studies (cf. Introduction). As the translator did not have to add van deze biografie after non-expression of the referent in the preceding sentences, the cohesive explicitness in the target text is clearly a personal preference inherent to the translation process:

Example (6) shows a simple reformulation (135 cases): pour s’engager dans cette voie is translated as initiatieven ter zake in Dutch (= des initiatives à ce propos / en la matière), which does not contain a demonstrative determiner, but it could also be literally translated as om deze weg in te slaan (with a demonstrative determiner). The translation by initiatieven ter zake is a stylistic choice on the translator’s part.

Example (7) is an example of reformulation by a relative phrase. The noun phrases een hersteltermijn and die termijn are replaced by the relative pronoun qui in the target language. However, both constructions are possible in French and Dutch:

In some cases, the entire noun phrase with a demonstrative determiner is translated by a personal pronoun or vice versa (31 cases) or by a demonstrative pronoun or vice versa (28 cases). This finding can be explained by two factors: first, explicitation of the personal pronoun or the demonstrative pronoun using a noun phrase with a demonstrative determiner in the target text (cf. explicitation hypothesis), as in examples (8) and (9), and second, the decision of the translator to replace the noun phrase by a personal or demonstrative pronoun in the broader context of the preceding text (cf. implicitation), as in examples (10) and (11):

However, when a noun phrase with a demonstrative determiner is replaced by an adverb or vice versa (66 cases), this is not always due to translator preference. In (12), for instance, the French – Dutch translator could have translated the French prepositional phrase including a demonstrative determiner (sur ce point) by a Dutch prepositional phrase with a demonstrative determiner (op dit punt) or by a pronominal adverb in Dutch (daarover). In (13), in contrast, the Dutch – French translator had to translate the Dutch pronominal adverb (hierover) as a prepositional phrase in French (à ce sujet), since these kinds of adverbs (e.g., hiermee, daartussenin, daartegen, daarvoor) do not exist in French.

In summary, examples (4) to (13) show that in cases of omission or reformulation of the entire noun phrase, the French or Dutch demonstrative determiner could also be expressed differently in the target language. In these cases, the French or Dutch demonstrative determiner does not seem to be structurally untranslatable.

In contrast, in the following examples, the translation shifts with respect to the demonstrative determiner are not a matter of translator preference. This is especially the case when another definite determiner (particularly the definite article) is involved in the target text.

In example (14), for instance, the addition of the adjective interne causes a determiner shift from the demonstrative determiner in the source text (cette note) to the definite article in the target text (de interne nota). While the addition of the adjective can be ascribed to translator preference and to translation universals, it also has an important impact on the determiner in the noun phrase (demonstrative determiner or definite article).

In examples (15) and (16), there is only a determiner shift. The definite article and the demonstrative determiner seem interchangeable in these sentences, but it should be determined whether any trends can be identified in the distribution differences between the French and Dutch demonstrative determiner systems, and whether these cases reflect a preference for either the demonstrative or the definite article.

3.2. Specific manifestations of translation universals within the noun phrase

Table 10, table 11 and table 12 show that syntagmatic and paradigmatic explicitation and implicitation within the target noun phrase may be mechanisms underlying determiner shifting in our corpus, particularly when the demonstrative determiner is translated by a definite article and vice versa. In these cases, syntagmatic and paradigmatic changes at the level of the head noun and its modifiers seem to be involved:

Table 10

Overview of changes within the noun phrase (French SL – Dutch TL)

Overview of changes within the noun phrase (French SL – Dutch TL)

-> See the list of tables

Table 11

Overview of changes within the noun phrase (Dutch SL – French TL)

Overview of changes within the noun phrase (Dutch SL – French TL)

-> See the list of tables

Table 12

Overview of changes within the noun phrase in the DPC sub-corpus

Overview of changes within the noun phrase in the DPC sub-corpus

-> See the list of tables

A first example of explicitation in the target text is the use of lexical expansion (syntagmatic change). This process of making the noun phrase longer, clearer or more complex is responsible for a large number of the instances of demonstrative shifting in the corpus (Chesterman 2001; Toury 2001). The insertion of an adjective (e.g., Rijselse) or a right-hand modifier (e.g., van Lombardsijde, de Rodenbach), for instance, causes a determiner shift, i.e., the demonstrative determiner is replaced by a definite article in the target text (17 cases). The type specifications of the nouns and its modifiers (Langacker 1991) in the target text are sufficiently rich to accept the definite article instead of the demonstrative determiner:

This determiner shift caused by lexical expansion also occurs in the opposite direction (5 cases): in the following examples, van Europa and specifieke are not translated or used in the target text, which causes the definite article in the source text to shift to the demonstrative determiner in the target text to compensate for the loss of type specification of the modifiers of the nouns in the target text. These cases involve syntagmatic implicitation rather than explicitation.

In the case of hyponyms (paradigmatic change), the emergence of the definite article or the zero article in the target text (French or Dutch translation), rather than the demonstrative determiner, seems to be significant (19 cases). In the following examples, the head nouns bouwsector, amygdala, Appels and Fondation are more strongly specified in the target text than in the source text, where we have secteur, structure, dernier and stichting (despite the presence of the modifier van openbaar nut). In these cases, paradigmatic explicitation of the head noun causes the demonstrative in the source text to be replaced by a definite article or a zero article in the target text to counterbalance the stronger type specification of the noun in the target text.

This determiner shift also occurs in the opposite direction, when a definite article and a hyponym in the source text are translated by a demonstrative determiner and a hyperonym with less semantic information in the target text (7 cases). In the following examples, the type specification of événement and team is less rich than for wereldrecordgebeuren and Doubistes, which has to be counterbalanced by the demonstrative determiner in the target text:

In summary, syntagmatic and paradigmatic explicitation of the noun phrase in the target text involving different type specifications of the noun and its modifiers may cause the demonstrative determiner in the source text to be replaced by a definite article (or, to a lesser extent, a zero article) in the target text. Conversely, syntagmatic and paradigmatic implicitation of the noun phrase in the target text may cause the definite article in the source text to be replaced by the demonstrative determiner in the target text:

Table 13

Syntagmatic and paradigmatic explicitation and implicitation involving demonstrative shifting

Syntagmatic and paradigmatic explicitation and implicitation involving demonstrative shifting

-> See the list of tables

3.3. Structural differences between the French and the Dutch demonstrative determiner

What happens if we only consider the cases of complete NP – NP correspondence with a determiner shift in the corpus, as in examples (3), (15) and (16)? Is there any other mechanism underlying this process? This question can only be addressed by examining possible structural differences between the French and Dutch demonstrative determiner systems.

Table 14, table 15 and table 16 show that NP changes at the determiner level mostly involve a shift between the demonstrative determiner and the definite article (60%). In the other corpus instances, the indefinite article, the possessive determiner and the zero article tend to be used. In what follows, we will focus on the shift from the demonstrative determiner to the definite article and vice versa.

Table 14

Overview of changes at the determiner level (French SL – Dutch TL)

Overview of changes at the determiner level (French SL – Dutch TL)

-> See the list of tables

Table 15

Overview of changes at the determiner level (Dutch SL – French TL)

Overview of changes at the determiner level (Dutch SL – French TL)

-> See the list of tables

Table 16

Overview of changes at the determiner level in the DPC sub-corpus

Overview of changes at the determiner level in the DPC sub-corpus

-> See the list of tables

Analysis of the corpus instances of complete NP – NP correspondence with a demonstrative shift to the definite article and vice versa reveals that, regardless of the direction between source and target language, significantly more French demonstratives are translated by the Dutch definite article than vice versa (2.23/1). The result reported in table 17 is only slightly lower than the result presented in table 6 (coefficient: 2.48/1):

Table 17

Relative frequency of the demonstrative – definite article and demonstrative – other structure pairs in the DPC sub-corpus

Relative frequency of the demonstrative – definite article and demonstrative – other structure pairs in the DPC sub-corpus

-> See the list of tables

A closer look at these data reveals a number of structural differences between the French and Dutch demonstrative determiner systems that cannot be ascribed to the translator preference and translation universals, nor to specific manifestations of translation universals in the noun phrase. These divergences can be explained by purely linguistic factors, such as fixed expressions (3.3.1.) and semantic differences in the referential force of the demonstrative determiner (3.3.2.).

3.3.1. Fixed expressions

The examples below illustrate lexical differences between French and Dutch involving the demonstrative determiner:

For instance, the Dutch time phrase in die tijd uses a demonstrative determiner, whereas the French translation à l’époque uses a definite article, although à cette époque-là also exists. In (28), the phrases à l’époque and in die tijd (just as indertijd = in de tijd van toen) are fixed expressions referring to a vague time period in the past. A modifier can be added after à l’époque, which is then no longer a fixed expression and indicates a specific point in time. In Dutch, a modifier can also be added after in de tijd, but not after in die tijd.

In example (29), à l’heure actuelle, another time expression, has been translated by op dit moment in the target text, which is not a simple case of implicitation of the adjective actuelle involving a demonstrative shift, but reflects a lexical difference between French and Dutch. First of all, à l’heure actuelle is a fixed expression mostly translated by a demonstrative determiner and without an adjective in Dutch (op dit moment). Alternatively, it can be translated as op het huidige moment or by op dit huidige moment. In contrast, the construction à cette heure actuelle does not exist in French. On the other hand, op dit (huidige) moment can be literally translated en ce moment (même), which has the same meaning as à l’heure actuelle. Finally, the phrase à cette heure, meaning nowadays as in (29), is not considered standard French, although it is used in Walloon,[3] in Louisiana Creole (Neumann-Holzschuh 1985), in Acadian French (Butler 1995) and in Quebec French[4] with variant spellings (asteur, asteure, asteûre, astheure, à c’t’heure).

Lexical differences involving demonstrative shifting also occur in phoric uses of the demonstrative determiner. For instance, in example (30), the fixed expression de klus klaren (22 500 Google hits)[5] referring to the responsibilities of the national health service in the preceding text, is more frequently used than deze klus klaren (2 970 Google hits), whereas in French mener cette mission à bien (2 950 Google hits) is more frequent than mener la mission à bien (265 Google hits). Note, however, that in this example the demonstrative determiner is possible in both languages.

3.3.2. Semantic differences

There are also important semantic differences in the referential force of the demonstrative determiner in French and in Dutch:

In its recognitial use,[6] (Diessel 1999; Gary-Prieur 1998; Himmelmann 1996; Jonasson 1998; Kleiber 2005; Kleiber and Sock 2006), the French demonstrative determiner (source language) is translated by the definite article in Dutch in 31.30% of cases. In all these cases, it is only possible to use a demonstrative determiner in French, whereas in Dutch both the demonstrative determiner and the definite article, combined with generalizing elements (such as on, toujours, en principe), are possible. In (31), for instance, the demonstrative determiner is not excluded in Dutch, but its strong referential (i.e., situational and/or textual) force makes it less suitable for recognitial use. Hence, the definite article is more appropriate. In French, in contrast, the demonstrative determiner is bleached (see e.g., De Mulder and Carlier 2006), making it appropriate for recognitial use. The definite article is excluded.

This fundamental semantic difference between French and Dutch also exists for some anaphoric uses (i.e., textual reference) of the demonstrative determiner. In the next example, the noun phrases cette collection and ces 956 pièces have been translated using the Dutch definite article in the target text (de verzameling and de 956 stukken). Again, the demonstrative determiner is not excluded in Dutch, but its strong spatio-temporal reference would seem to exceed the requirements of textual cohesion. On the other hand, in French, the textual recuperation in the form of a noun phrase with a definite article would be too weak.

This is also the case in the next example: both the Dutch definite article (de bekende Antwerpse advocaat) and the French demonstrative determiner (cet avocat bien connu à Anvers) have normal anaphoric force, whereas the Dutch demonstrative determiner (deze bekende Antwerpse advocaat) has very strong anaphoric force and ensures the effectiveness of the anaphoric process. The anaphoric force of the French definite article (l’avocat bien connu à Anvers), on the other hand, is too weak.

The semantic difference between French and Dutch with respect to the anaphoric use of the demonstrative also affects cleft and presentative sentences, as illustrated in examples (34) and (35). The cleft construction c’est (X) ce Y que … has to be translated as het is die Y die / waarin / waardoor … in Dutch, which demonstrates the bleaching of the demonstrative pronoun ce in c’est. A similar observation can be made with regard to presentative sentences: in the construction c’est (X) ce Y que …, the demonstrative pronoun ce in c’est is bleached and accepts a demonstrative determiner, whereas the demonstrative pronoun dat in dat is is too strong in Dutch and does not accept a demonstrative determiner (*dat is die Y die / waarin / waardoor…). In the latter construction, the definite article is used in Dutch (dat is de Y die / waarin / waardoor…).

Finally, some time phrases in rhematic position, expressing anteriority with respect to the moment of speech, such as ces derniers temps (36) and ces vingt dernières années (37), are always translated by a definite article in Dutch (de jongste tijd, de laatste twintig jaar) rather than by a demonstrative determiner. In contrast to the French demonstrative, the Dutch demonstrative has very strong referential force and hence does not accept a numeral or the adjective laatste in the same time phrase (*deze laatste (twintig) jaar).

4. Conclusion

The aim of this paper was to study the translation of the demonstrative determiner in French and Dutch in parallel corpora, in order to determine which types of translation shifts occur systematically and what the underlying mechanisms and semantic effects of this process are. Analysis based on a sub-corpus of the Dutch Parallel Corpus showed that translational shifts with respect to the French and Dutch demonstrative determiner occur at three different levels: at the level of the noun phrase, within the noun phrase and at the level of the determiner. Table 18 and table 19 give a detailed overview of these findings:

Table 18

Overview of translational shifts in the French SL corpus and the Dutch TL corpus

Overview of translational shifts in the French SL corpus and the Dutch TL corpus

-> See the list of tables

Table 19

Overview of translational shifts in the Dutch SL corpus and the French TL corpus

Overview of translational shifts in the Dutch SL corpus and the French TL corpus

-> See the list of tables

First, the analysis suggests that – with respect to the French and Dutch demonstrative – translator preference is mainly involved in instances of omission, addition or reformulation of the noun phrase (30.73% of all instances in the French SL – Dutch TL corpus, 83.03% of these instances involving demonstrative shifting, and 27.47% of all the cases in the Dutch SL – French TL corpus, 79.08% of these cases involving demonstrative shifting). These instances illustrate the inherent variability of the translation process rather than a divergence between the Dutch and French determiner system, as these noun phrases containing a French or Dutch demonstrative determiner are clearly not ‘untranslatable’ and could easily be expressed differently.

Second, in order to verify whether specific manifestations of translation universals within the noun phrase could explain other translation shifts with respect to the demonstrative, syntagmatic and paradigmatic changes in the noun phrase were examined (11.55% of all instances in the French SL – Dutch TL corpus, 77.78% of these cases involving demonstrative shifting, and 6.10% of all instances in the Dutch SL – French TL corpus, 67.65% of these cases involving demonstrative shifting). Findings show that the interaction between the demonstrative determiner and the definite article is strongly dependent on the type specification of the noun and its modifiers. As was illustrated above, syntagmatic and paradigmatic explicitation of the noun phrase in the target text involving different type specifications of the noun and its modifiers may cause a demonstrative determiner in the source text to be translated by a definite article in the target text. Conversely, syntagmatic and paradigmatic implicitation of the noun phrase in the target text may cause a definite article in the source text to be translated by a demonstrative determiner in the target text.

Third, analysis of instances of complete NP – NP correspondence involving only a determiner shift (21.42% of all instances in the French SL – Dutch TL corpus, and 21.36% of all instances in the Dutch SL – French TL corpus) revealed structural divergences between the French and Dutch demonstrative determiner systems. More precisely, French and Dutch were found to use different fixed expressions at the level of the temporal (e.g., à l’époque vs in die tijd, à l’heure actuelle vs op dit (huidige) moment) and the textual (e.g., mener cette mission à bien vs de klus klaren) uses of the demonstrative determiner in French and Dutch. The findings also revealed considerable semantic differences in the referential force of the demonstrative determiner in French and Dutch. The French demonstrative is more bleached than the Dutch demonstrative, and is more frequently used in some specific contexts where Dutch would prefer the definite article (e.g., recognitial use, indirect anaphors, specific temporal uses).

Finally, this paper tries to demonstrate the usefulness of a detailed parallel corpus study, which clearly distinguishes between changes at different levels, in accounting for divergent translations of the demonstrative determiner in different languages. To this end, several types of explanation, drawn from various fields (such as translation studies and contrastive linguistics), must be considered.