‘Loss’ or ‘lost’ in translation: a contrastive genre study of original and localised non-profit US websites
Miguel A. Jiménez-Crespo, Rutgers University, The State University of New Jersey
ABSTRACT
The notion of ‘translation loss’ has been at the core of numerous publications and theorisations in Translation Studies (Nord 2011). This paper revisits this notion in the light of new dynamic digital genres with open hypertextual structures that are constantly enlarged and updated. The localisation of these websites, understood as a process constrained by finite time and financial resources, normally entails deciding or prioritising which sections or blocks will appear in the localisation and which will not. These strategies have an impact in the configuration of the localised websites and their potential impact on the communities they address. The present genre-based study analyses from a descriptive perspective a representative corpus of the ‘non-profit website’ genre in the United States and contrasts non-translated original websites to their Spanish localised versions. After a descriptive study of this digital genre, a contrastive superstructural study shows the differences between these two distinct textual populations in terms of ‘localisation loss,’ or the chances of sections of the source website not being localised. This analysis sheds some light on the perspective of producers and translation commissioners on the genre, and on the potential differences between non-translated and translated digital genres.
KEYWORDS
Web localisation, technical translation, translation loss, digital genres, comparable corpus, non-profit website.
1. Introduction
Since the introduction in Translation Studies of the notion of genre from Discourse Analysis and English for Specific purposes (Swales 1990; Bhatia 1993; Paltridge 1997), descriptive and contrastive studies of the most commonly translated genres have been at the core of the discipline. Studies based on the notion of textual genre depart from the premise that textual analysis prior to translation is paramount to the practice and teaching of translation (Nord 1997; García Izquierdo 2000; Gamero 2001; Hurtado 2001), as well as a necessary theoretical and methodological construct. Hurtado considers genre a key notion:
The description and classification of textual genres is essential in Descriptive Translation Studies, as it allows to better know the modalities and translation types, but also for the training in translation (Hurtado 2001: 505).1
The introduction of corpus based methodologies in translation research since the 90’s (e.g. Baker 1993, 1995; Laviosa 1998) increased exponentially corpus-based genre studies, mostly from a contrastive perspective (Baker 2004: 1969). Normally, these studies focus on the description of and convergences between similar genres in different sociocultural contexts serving the same purpose and communicative situation, such as purchase contracts, powers of attorney, research papers, etc. Studies have also focused on the differences between non-translated and translated genres in order to research the impact of the process on translated products (Jiménez-Crespo 2008, 2009a, 2009b, 2010a, 2011a). The differences the same genres might show between cultures are due mostly to the randomness in which genre conventions are established in each sociocultural context (Lewis 1969; Nord 1991, 1997), as well as the impact of the sociocultural context in their emergence and evolution (Hatim and Mason 1990; Berkenkotter and Hucking 1995). As a result, genres serving the same purpose in different cultures show distinct conventions and lexical, syntactic, pragmatic or discourse features.
In this context, localisation as a new translation modality has been dealing with the unpredictable nature of the technological revolution, in which digital genres emerge and consolidate in unpredictable ways (Santini 2007). The description and analysis of these digital genres has attracted the attention of a small number of researchers, focusing mostly on corporate (i.e. Jiménez Crespo 2010a, 2010b, 2009a, 2009b, 2008; Diéguez 2008; Bouffard and Craignon 2006; Bolaños 2003), institutional (Fernandez Costales 2008), social networking (O´Hagan forthcoming; Jiménez-Crespo 2011b), or tourist websites (Jiangbo and Ying 2010; Capelli 2008; Pierini 2007). However, no study has currently focused on the localisation of non-profit websites from an intracultural perspective, that is, non-profit organisations that operate within a single society in which one or more communities speak a language other than the official language(s) of the region or country. The study of this digital genre is therefore long overdue. The purpose of this paper are twofold, (1) to describe this emerging genre that might potentially soon have the highest rates of localisation, and (2) to analyse, in the context of a time and resource constrained process such as localisation (Dunne 2006; Wright 2006), what sections or communicative blocks are selectively chosen to be localised into Spanish. This contrastive study breaks new ground in the description of this ‘localised genre’ as a distinct textual population, and it represents a new approach to the notion of ‘translation loss.’ Spanish was chosen as it is the most spoken language in the US after English, with over 50 million speakers according to the 2010 US census.
The basic premise supporting this genre-based study is that hypertexts are by nature open structures, where content or sections can be constantly added and/or changed (Landow 1992; Janoschka 2003). However, the commission2 of the web localisation process, again in the context of a constrained process, determines which content might be localized and which sections or content might not. Therefore, the description not only of the source genre, but also the translated versions, sheds some light onto whether localised genres can be considered “alternative genres on their own” or “localized genres” (Jiménez Crespo 2009a). It can also reveal whether their genre description, i.e. textual function, intratextual configuration, communicative situation, etc., matches that of non-translated similar websites. This phenomenon has clear social implications given the impact these websites might have on immigrant and minority communities.
In this study, the working notion of ‘loss’ does not refer to the much debated issue of ‘translation loss’ in literary translation research (Bassnett 2002; Classe 2000). Literary approaches to translation understand the notion of translation loss in reference to some sort of equivalence to a source text with a superior status to the translation itself (Nord 2011), and normally lament certain necessary loss with mostly negative connotations. It cannot also be associated with the notion of ‘omission’ (Vazquez Ayora 1977; Delisle 1993) or ‘dissolution’ (Vinay and Dalbernet 1958) as a valid and conscious translation strategy, or with the notion of ‘omission’ as an error in translation and interpreting assessment. In the specific case of localisation strategies, translation or ‘localisation loss’ from a superstructural and macrostructural perspective can be understood as a specific strategy to deal with the dynamic and open nature of hypertexts in the context of time and resource constrained processes. Thus, the commission of the process might request a static version of the website without an actual localisation of dynamic sections such as press releases or calendars. The more resources that commissioners are able to allocate to web localisation, the higher the chance that more sections or complete localisation will take place. This view of the localisation process corresponds partially to Pym’s notion of ‘partial localization,’ understood as “a process in which not all of the user-visible language is translated, usually to save costs when working into a small locale” (Pym 2010: 123). Consequently, this entails a completely different concept to those underlining notions of loss throughout the evolution of translation as an object of scholarly study. This study therefore adopts a descriptive approach to describe the strategies of non-profits to deal with unavoidable constraints.
The paper is structured in the following fashion. After a detailed description of the corpus-based methodology used for the corpus compilation, the digital genre ‘non-profit’ website is described. This is followed by a contrastive superstructural study of the compiled original and localised websites in order to elucidate the prevailing strategies in the non-profit sector and to describe the differences between the original and translated versions of this digital genre in the US.
2. Methodology for corpus compilation and textual analysis
Corpus linguistic approaches to translation research have been extremely productive during the last couple of decades (Laviosa 2010). A carefully constructed parallel corpus can be a source of conceptual, terminological and linguistic information as well as a quantitative base in order to research translation processes and products (Olohan 2004). In general, a corpus can be defined as “a collection of authentic texts held in electronic form and assembled according to specific design criteria” (Laviosa 2010: 80). Among different corpus types, a parallel corpus can be defined as a corpus “that contain a series of source texts aligned with their corresponding translations” (Malmkjaer 1998: 539), either into one or multiple languages. Some examples of multilingual parallel corpora include the Europarl corpus of European Parliament procedures (European Parliament) the JRC-Acquis of European legislation or the Open Subtitles corpus of movie subtitles.
The theoretical design in the creation of any textual corpus is extremely important and it should precede its design or compilation (Biber 1993:13). Among the several issues to consider, such as text typology, genre or standardisation, representativeness is the most significant. According to Biber “a corpus is put together in a principled way so as to be representative of a larger textual population, in order to make it possible to generalize findings concerning that population” (Biber 1993: 243). It is therefore necessary that the corpus includes “the full range of variability in a population” (ibid). In order to compile a representative corpus of the textual population targeted, —the websites of US non-profits that serve disadvantaged groups—the directory of the US Internal Revenue Service for non-profit organisations was used. This website offers the possibility of searching non-profits by city and state, as well as a word search function in order to find NGO’s using lexical items.
Figure 1. Search Engine by the US Internal Revenue Services to locate non-profit organisations.
The study was circumscribed to four states in the four areas with the highest concentrations of Hispanics in the US, the Mid and SouthWest (Texas and California) and the South and North East (Florida and New Jersey).3 Additionally, and given that the study focuses on non-profits that serve disadvantaged groups, four areas were selected: legal services, healthcare, work and nutrition/food. For each area, one or more lexical units were used in order to probe the IRS non-profit search engine. The lexical units selected were ‘legal’ for legal service non-profits, ‘medical,’ ‘health’ and ‘cancer’ for healthcare, ‘work’ and ‘employment’ for job related non-profits, and finally, ‘food’ and ‘meals’ for nutrition non-profit websites. These lexical units or terms were selected after an initial exploratory study in order to provide a comprehensive analysis and avoid the noise that would result from retrieving non-profits such as medical professional organisations that do not have as a goal assisting disadvantaged or minority groups. The search using these lexical units in these four states yielded a total of 1890 non-profit organisations. Of that amount, only 943 offered a website in English. In the search for localised websites, it was identified that only 6.68% offered some content in Spanish, while merely 2.54% offered a fully functional mirror website localised into Spanish. A working notion for the search of localised versions was ‘localisation level,’ or the extent to which websites are localised. This level varies according to the translation or commission, from a simple page with the contact information to a fully localised site. Other forms of offering content translated into the target languages that do not involve strictly speaking web localisation were identified, such as uploading translation in pdf format to the main English site, or providing a link to Machine Translation engines, such as Google Translate or Microsoft.
The corpus was compiled during April 2011, and it includes the 63 source websites and their respective localisations into Spanish. Once the corpus was compiled, it was analysed from a genre perspective with a previously established methodology used in corporate websites (Jiménez 2008, 2009, 2010, 2011a). The main premise behind the genre analysis, as with any other printed genre, is that genre structure is formed by a series of textual segments, the communicative blocks and sections in hierarchical order. In different studies from a Discourse Analysis perspective these textual segments have been called sections>moves>steps>substeps (Swales 1990), triad>keys (Paltridge 1997) or communicative blocks>communicative sections>significant units>significant subunits (Gamero 2001). Digital genres differ from printed genres in that they show hierarchical or linear structures, while digital ones are multilinear, either axial or networked (Landow 1992). Communicative blocks are typical of each genre and in web hypertexts, they can be visually represented in the different links that produce the mutlilinear superstructure (Askehave and Nielsen 2005). These textual blocks make up the different parts of a global text, the hypertext, and each of them conveys a specific function in the global multifunctional text (Swales 1990). As an example, the communicative section ‘Donate’ in a non-profit website expresses an exhortative or appellative function since intends to both persuade users to collaborate with the organisation and to establish a relationship of trust and confidence with the potential donor or volunteer.
As previously mentioned, genre structure is culturally bound (Shreve 2006), and differences are found not only between similar genres in source and target cultures, but also between non-translated and translated ones. As an example of the cross cultural differences between original and translated digital genres, in a study by Jiménez-Crespo (2011a) the section ‘Privacy Policy’ or ‘Política de Privacidad’ appeared in 42.4% of US websites localised into Spanish, while in the corporate website genre originally produced in Spain, only 13.6% showed this term. Spanish corporate websites usually include a ‘Legal’ block under the term ‘Avisos Legales,’ and this block usually includes any privacy legal provisions. This aspect is indicative of cultural differences in the conventionalisation of genre structure, and it points at the mere reproduction of the culture bound source textual structure (Jiménez-Crespo 2011a). At the same time, each block might be divided into communicative sections, and they also represent a specific function inside each communicative block (Gamero 2001). The section ‘Board’ inside the block ‘People’ in a non-profit site shows the user who is directing the non-profit, and it is a specific section that appears in 52.5% of US nonprofit websites according to this study. The methodology used is based on the assumption that each webpage entails a distinct storage and information unit (Brink et al. 2002; Nielsen and Loranger 2006), even when different blocks might be combined in one single page. The creation of a different unit of information or webpage is indicative of the existence of a communicative section in the digital genre (Askehave and Nielsen 2005). Nevertheless, and despite the fact that the methodology used only accounts for a different communicative block or section if a separate webpage exists, this does not mean that this content might appear within other pages. For example, ‘About us’ pages might include contact or location information, even when they are identified as different communicative sections in this study. However, developers or text producers did not deem this specific information or content as relevant enough to be included in a separate distinct page with an entry in the navigation menu. The contrastive analysis in this study therefore accounts for separate webpages for each content.
The methodology to analyse and identify the prototypical superstructure of the genre therefore involves assigning each lexical unit in navigation menus in source English and Spanish websites to one of the communicative blocks or sections identified in a preliminary exploratory study. On a second level, many of the identified communicative blocks were assigned to a subdivision or communicative section within a larger block: for example, fifteen different sections were identified in the block ‘About us,’ such as ‘Location,’ ‘Foundation,’ ‘Mission,’ ‘History,’ etc.
2.1 Parameters in the description of genres and digital genres
Genres can be defined from a discourse perspective as communicative acts that express themselves through conventionalised forms of texts, therefore increasing the communicative efficiency in a recurring particular social situation. Hatim and Mason (1990: 69), from a discourse perspective defined genre as “conventionalised forms of texts which reflect the functions and goals involved in particular social occasions as well as the purposes of the participants in them.”
This definition combines formal aspects, such a prototypical structure, social and cultural aspects, since genres are determined by a specific culture and social occasion, and cognitive ones, since genre represents the purposes and expectations of both the sender and the receiver of the text. Several genre description models for printed genres have been used in Translation Studies, and they take into consideration several aspects, such as conventions, textual functions, the communicative situation, the social and cultural context and intratextual elements. Among the intratextual elements, textual structure plays an essential role in genre description. According to Swales “exemplars of a genre exhibit various patterns of similarity in terms of structure, style, content and intended audience” (Swales 1990: 58). The model of genre description by Gamero (2001) focused on technical translations and was adopted for this study. The researcher, from an interdisciplinary perspective, uses the following parameters to describe genres: textual function, conventional items, communicative situation, sociocultural context and intratextual elements. Additionally, in the specific case of digital genres the functionality, understood as the possibilities of interactivity and interaction, need to be added to the characterisation model of these genres. This is due to the fact that since the emergence of digital genres, functionality has been the main force behind the evolution (Shepherd and Watters 1998; Crownston and Williams 1999; Santini 2007). Genres are in constant evolution (Miller 1984), and the evolving functionalities in the new medium have been the reason behind both the emergence of a number of new genres, as well as the adaptation of existing genres to the web. This would be the case of printed vs. online newspapers, and their further evolution into RSS news feeds and the recently emerged online news aggregators.
Figure 2. Evolution of digital genres from printed to online forms according to Shepherd and Watters (1998).
From a digital or cybergenre perspective, the most widely used description model is that proposed by Shepherd and Watters (1997). The researchers propose describing digital genres using three parameters: (1) content, (2) form and (3) functionality. Following this model, they described three primary genres, the homepage, brochures, resources, catalogues and search engines. As an example, the genre homepage has as content primarily information about the company-person or institution, its form includes introduction, hierarchical images and animated images, and its functionality tend to include browsing and e-mails. It is easy to see how digital genres have dramatically evolved since this digital genre classification model, but nevertheless, a number of studies have used the model in digital genre studies, mostly from a linguistic (Schmid-Isler 2000; Fernandez 2003; Alaman 2003) or Translation Studies perspective (Bolaños 2003; Rennau 2004). For the purposes of this paper, the genre description model of Gamero (2001) is adopted as it allows us to take into account in depth linguistic and discursive factors that the above mentioned cybergenre model cannot.
Once the models of genre and cybergenre description have been briefly described, the following section presents the digital genre under study, the intracultural ‘non-profit’ website.
3. Description of the genre ‘non-profit’ website
The description of the genre will follow the above mentioned model with the incorporation of functionality as the main driving force behind the emergence and evolution of all digital genres (Shepherd and Watters 1998; Santini 2007).
3.1 Textual function
As with any other text, the producers of a non-profit website have an intention behind it (Beaudegrande and Dressler 1981). This shapes the potential textual function, an essential component of all genre description models (Swales 2004; Göpferich 1995; Bhatia 1993). Textual function is not inherent to any text, but is assigned by the user or receiver in any given situation (Nord 1997). Depending on the discipline or approach, textual function classifications have different denominations but, all proposals include three basic types: exhortative or persuasive, expositive and argumentative. In contemporary genre theory, it is assumed that text does not express just one function, but rather, they are multifunctional, with one or two primary functions and a potential secondary function (Hatim and Mason 1997; Göpferich 1995; Gamero 2001).
In principle, studies have shown that most websites normally have a primary exhortative or persuasive function with a secondary expository one. The reason behind this argument is that in the massive WWW world, attention spans are shorter and normally users do not read websites, they scan them for the information they need, and if they do not find the required information quickly, they move on to the next site (Nielsen and Loranger 2006; Brink et al. 2003). However, it can be argued that in the intracultural non-profit genre, as opposed to the international non-profit website genre, the main purpose would be to provide information to potential users, with a secondary exhortative function to show a professional image, in order to both create a good image for potential users and to attract potential donations and volunteers to the cause of the non-profit. This would be different from international non-profit organisations (Red Cross, Amnesty International) in which the potential receivers of services would not normally have access to the website, since the site is mostly addressed at donors, volunteers and international press. Thus, the main difference between intracultural and international non-profit website genre could be precisely the primary and secondary functions. The three most frequent sections, ‘Description-services’ (80%), ‘About us’ (77.5%) and ‘Contact us’ (77.5%) demonstrate that the primary function in the genre under study is mostly expositive. Meanwhile, the sections mostly addressed at potential donors and volunteers appear in 60 % and 42.5% of sites respectively.
3.2 lements of the communicative situation
As with all websites, the communicative situation can be described within the Internet Mass Communication model (Janoschka 2003). This model combines elements of a Mass Communication Model, together with interpersonal communication. This Communication Model is also characterised by a triple communicative process in which the senders communicate with the audience, and the website communicates with the users in interactive sections or if the information provided in forms is filled out incorrectly. At the same time, website communicates with the wider WWW, that is, the description and metadata that helps search engines categorise and retrieve the information. In the case of non-profit websites, the organisation disseminates both static and dynamic content to the wider WWW community, in a communication flow mostly from the organisation towards its audience, although normally websites include interactivity options, such as forms, subscriptions to newsletters, etc. Static information refers to that information that does not change over time, such as contact information or the description of the non-profit. The dynamic content includes all the sections that might be regularly updated, such as calendar, press, events or newsletters. As with all websites, the audiences can be divided between (1) primary audience, (2) supervisory audience, and (3) peripheral audience (Jeney 2007). The website specifically addresses the primary audience, and in the case of non-profits this audience is comprised of both the recipients of services by non-profits as well as the potential donors and volunteers. Donors and volunteers are targeted through the communicative block ‘Donate’ seen in 60% of the websites analysed. The supervisory audience refers to the commissioners of the website, that is, both the organisation that commissioned the website and/or the translation agency that might have been in charge of the localisation process. Finally, the peripheral audience would be all those visitors that might not be directly targeted when the website was created, but that might nevertheless visit the site. Translators or localisers are never the primary audience in any mediation process, but rather, they are normally part of the peripheral audience of any text as they are not the intended receivers of any potential text to translate (Muñoz Martín 1995).
As section 4 will show, the communicative profile of the localised websites diverges from that of the source website to some extent. We see that the primary audience is different due to the fact that the sections ‘Donate’ and ‘Volunteer’ appear in 60% of original sites and 42.5% of localised ones, while they only appear in only 35% and 15% of localised sites, respectively. In the case of ‘Volunteer,’ the chances of this section being lost in localisation amounts to 64.7%. This shows to some extent that the primary audience of the site in an intracultural situation, as reflected by the conscious decisions of the commissioners or initiators of the localisation process, is a linguistic or ethnic minority group within one society that is different from the mainstream audiences of the source website.
3.3 Influence of the sociocultural factor
All genres are contextualised in a specific sociocultural environment and they develop due to a repeated communicative situation or need (Hatim and Mason 1990; Berenkotter and Hucking 1995). Even when most digital genres appear in most Western nations, there is still the possibility of genres being culture specific, such as the Japanese haiku. The sociocultural contexts impacts genres in three ways: (1) Conventions are culture specific and cannot be directly transferred (Nord 1997); (2) they are key to the emergence of new genres, without any specific communicative need within a society, genres will not develop; and (3) it impacts the evolution of any genre as it adapts to any changes within that society. As an example, in monolingual societies without immigrant communities, localised web non-profit websites might not exist. Similarly, in most Western nations with socialised medicine, non-profit websites of health care organisations that serve the poor might not exist, and in this case, it might be a culture-specific genre in the United States or other countries without single payer health systems.
3.4 Functionality
Interactivity is considered the main defining characteristic of digital genres as opposed to printed ones, and it was the foundation for the first classification of digital genres or cybergenres (Crownston and Williams 1997). Functionality was also the main engine of evolution of digital genres in its early days (Shepherd and Watters 1998; Kennedy and Shepherd 20005). Nowadays, the interaction possibilities are highly conventionalised, as users expect for example, a search function on large websites, and subsequently, it is expected that the search results will include links to the webpages match the search criteria. The non-profit websites compiled are characterised for low levels of functionality. For example, only 37.5% include a search function and only 32.5% of them a sitemap.
3.5 Conventional items
Conventions are a key element in the description of any genre as they assist in facilitating communication and comprehension, as well as functionality in the WWW (Jiménez Crespo 2009a). Genre conventions are essential in the translation process as they differ between cultures and language (Nord 1997; Kussmaul 1997), and therefore, translators have to be aware of conventional items in the source and target genres, making sure they appear in instrumental (Nord 1997) or covert (House 1997) translation types. This is the case of all digital genres, as they all are supposed to represent a functional text in the target system, without users being aware that the text they are using and interacting with is necessarily a translation. However, in cases of intracultural mediation such as the one under study, translators might not have a culturally-bound model upon which to compare or extract any potential target conventions. This is even more so in the case of Spanish and the rise of the so-called US-Spanish. Conventions in genres are displayed at the intratextual level, mostly in terminology, phraseology, syntax, and most importantly, at the superstructural level. As an example of the conventions in source websites, the lexical units used to direct users to the Spanish version of the site was analysed. Figure 3 shows that the most frequent lexical unit to direct users to the Spanish version is ‘Español’ or ‘En español,’ with 6.06% of use of ‘Espanol,’ [sic] without the Spanish diacritical sign for the “ñ,” a major typographic error in this language. It can also be observed the 9.09% of websites do not include any type of link to the Spanish content, and this Spanish localised content cannot normally found until users surf deeper into the website structure.
Figure 3. Percentage of use of lexical units indicating the link to the localised version of the website.
3.6 Intratextual elements
Normally, genre description models favor extratextual elements over intratextual elements. The conventions are mostly expressed at the intratextual level, such as specific terminology, phraseology or textual structures. Most studies focus normally on one or two of intratextual elements when describing any genre, such as superstructure, speech acts, syntax, lexis, tone, thematic progression, cohesion, metacommunicative elements etc. However, not all these elements might be equally relevant to all genres. Among these elements, researchers agree that superstructure is the most important of all given that is generally conventionalised in most genres, and that it provides the necessary cohesion to them (Gamero 2001; Göpferich 1995). Additionally, the top—down analytical progression that characterises most contrastive or descriptive studies starts at the superstructural level, then progresses to lower levels focused on microstructures. Given the importance of superstructural analysis in genre studies, the following section analyses this aspect in the cybergenre under study.
3.6.1 Superstructural analysis of the digital genre ‘non-profit’ website
The analysis of the textual population compiled in the corpus through a structural analysis of navigation menus and sitemaps identified nine communicative blocks in the prototypical structure of English non-profit websites: (a) Homepage, (b) Contact us, (c) About us, (d) Legal, (e) Press- News, (f) Jobs-Careers, (g) Donate and (h) Interaction and (i) Resources. Within these nine main blocks, thirty communicative sections were identified, mostly in the blocks ‘About us,’ ‘Legal,’ ‘Multimedia’ and ‘Interaction.’ Table 2 shows the structural analysis of the nonprofit website in the US.
Communicaitve |
Communicative sections |
% in Source English Websites |
A.Home |
65% |
|
B. Contact us |
77.5% |
|
C. About us |
77.5% |
|
C.1. Description-Services |
87.5% |
|
C.2.1. People>Board |
52.5% |
|
C.3. Mission |
50% |
|
C.2.2. People>Staff |
37.5% |
|
C.4. Location |
35% |
|
C.5. Calendar |
25% |
|
C.6. Foundation |
22.5% |
|
C.7. History |
22.5% |
|
C.8. Testimonial |
17.5% |
|
C.9.1. Multimedia>Photos |
17.5% |
|
C.10. Sponsors |
17.5% |
|
C.9.2. Multimedia>Videos |
15% |
|
C.11. Finance |
10% |
|
C.12. FAQ |
7.5% |
|
C.13. Feedback |
7.5% |
|
C.14. Accesibility |
7.5% |
|
C.9.3. Multimedia>Audio |
2.5% |
|
C.15. Brochure |
0% |
|
D. Legal |
5% |
|
D.1. Terms of use |
25% |
|
D.2. Privacy Policy |
32.5% |
|
D.3. Disclaimer |
5% |
|
D.4. Equal Opportunity |
5% |
|
E. Press |
72.5% |
|
E.1. Newsletter |
7.5% |
|
E.2. Events |
22.5% |
|
F. Jobs/Career |
45% |
|
G.Donate |
60% |
|
G.1.Volunteer |
42.5% |
|
H.1. Interactivity>Site Map |
32.5% |
|
H.2. Interactivity>Search |
37.5% |
|
H.3. Interactivy>Links |
22.5% |
|
I. Resources |
35% |
|
I.1. Publications |
22.5% |
Table 1. Original non-profit website superstructure.
The superstructure of non-profit websites therefore resembles other kinds of homepages, such as institutional or corporate websites (Jiménez-Crespo 2009a, 2011a), in the main communicative blocks, with ‘Donate-Volunteer’ being one of the main differences in the structure of the sites as opposed to other homepage genres. However, in order to identify the conventional prototypical superstructure of this genre, any given block has to appear in a percentage of all examples of the genre, ranging from 90% (Fernández Sánchez 2004: 256; Nielsen 2004), to 70% (Hoffman 1988: 154; cf. Fernández Sánchez, 2004), or 50%. Blocks or sections with lower frequencies are considered as ‘occasional’ (Gamero 2001).
If the minimum 50% rate is applied to this genre, the prototypical superstructure of this genre would include the communicative blocks and sections shown in Figure 4:
Figure 4. Prototypical superstructure of the digital genre ‘intracultural non-profit’ website.
This prototypical structure analysis shows that the goal of this website genre is to provide information about the organisation services, press releases, mission or board, to provide contact information for users or donors, to and offer information about how to do donations.
Once the genre and its prototypical superstructure have been described, the next section of the study focuses on the contrastive differences between the original or non-translated English websites and their localisations into Spanish.
4. Lost in localisation: contrastive study of textual structures between source and localised websites
In order to perform the contrastive superstructural analysis of the source and localised genre, the superstructure of the 63 websites included in the corpus were analysed following the previously described methodology. Figure 5 shows the results of the contrastive study in which the blue lines indicate the frequency of use of that block or section in source English websites, while the red lines are indicative of the percentage of use in the localised versions of these same sites. It can be clearly observed that for all sections or blocks, the percentage of use in localised sites is considerably lower than that of source sites, and therefore, this figure graphically shows the considerable loss that source content suffers in the localised version.
Figure 5. Contrastive superstructural analysis of source and localised websites
First of all it is of interest to note that almost all blocks and sections appear in both textual populations under study. Only the block ‘Equal Opportunity’ and ‘Disclaimer,’ both in the legal section, appear with low frequency exclusively in original non-translated websites. Similarly, only one communicative block appears exclusively in the localised section, the block ‘Brochures.’ Normally, the upload of translated brochures in pdf format to any non-translated website entails the first strategy in disseminating information in languages other than English. In the case of US non-profits, 0.84% of English websites use this type of strategy.
If the chances of all the blocks and sections not appearing in the localised Spanish versions are compounded, the overall chances of sections of the source website not being localised are 44.18%, that is, on average, almost half the sections of any source website are not localised into Spanish. If this same calculation is performed exclusively on the conventional or prototypical sections of the non-profit website as described in Figure 4, the chances of a section not appearing in the localised version goes down to 34.07%. This could be expected, in part, because the more conventional and therefore important a section is, the higher chances it has of appearing in the localised version.
If the prototypical profile of localised websites is analysed, it can be observed that those sections with over 50% frequency would be only three: C.1. Description-Services (62.5%), C. About us (60%) and B. Contact us (52.5%). It can therefore be observed that from the eight sections in original websites with over 50% frequency of use, the percentage goes down to three in the prototypical profile of localised sites. However, it should be noted that those three are also the most frequently used sections in original sites.
The next analysis concentrates on the likelihood of any section in the source websites not appearing in the localised version. The variable for this analysis would be ‘localisation loss’ (LLoss), understood as the percentual chances of any section not appearing in the localisation relative to the percentage of use of that section in both original and localised sites. As an example, the section ‘Jobs-Careers’ appears in 45% of original sites and 20% of localised websites. The value of the variable LLoss would be 55.56%, as these are the chances that this section will not appear or will be lost in the localisation process. Figure 6 shows an analysis of the variable LLoss for the eight prototypical sections in source websites.
Figure 6. Possibilities of not appearing in the Spanish localised version for the eight conventional (frequency>50%) sections in the genre ‘intracultural non-profit website.’
It can be observed that the section ‘Press,’ a dynamic section that requires constant updating, has a higher chance of not appearing in the localisation (LLoss= 48.28%). This could be expected, since in the constrained localisation process, any dynamic section would be avoided as they would need a constant investment in localisation. Additionally, press releases from the organisation or with the organisation as a main topic might not be of equal interest to users of services and potential donors alike. This is followed by the sections ‘People>Board’ (LLoss=47.62%) and ‘Donate’ (LLoss=41.67%). This could account for the differences in the communicative situation between the original websites and localised sections described in section 3. The disadvantaged group addressed by the localised site is not the same as the collective of users that the section ‘donate’ addresses.
Figure 7. Possibilities of all blocks and section in the original ‘intracultural non-profit website’ of not appearing in the Spanish localised version.
The overall contrastive study according to the variable LLoss shows a complete picture of the sections that are more likely not to appear in localised versions. Figure 7 shows this analysis. The two sections with 100% percentages for this variables, ‘Equal opportunity’ and ‘Disclaimer’ represent two minor sections that appear just in 5% of original websites, and that do not appear in any localised version. This contrastive table shows in part that localised sites show lower levels of interactivity and of currency if compared to the source websites, therefore pointing to the more static nature of the localised versions. All interactivity sections, ‘Sitemap’ (69.23%), ‘Search’ (67.67%) and ‘Links’ (60%) appear among the sections with higher LLoss values. Dynamic sections that require updating, such as ‘Calendar’ (70%), ‘Events’ (55.56%), ‘Jobs’ (55.56%) or ‘Press’ (48.28%) also show considerably lower levels in the LLoss variable, and therefore, this proves a more static nature in the localised genre.
5. Conclusions
This study was initiated with the goals of describing the digital genre non-profit website in the US, both in its source and in its localised form. It has been observed that the open nature of hypertexts unavoidably leads to a localisation process with a certain degree of loss. This loss corresponds to the notion of ‘partial localisation’ (Pym 2010) and it might vary according to the digital genre or to the goals of the localisation process. The understudied genre ‘non-profit website’ in the United States was chosen because the United States offers more non-profits with a wider range of social services than any other Western Nation. In this sense, this genre clearly fulfills a social function in a country with an ever increasing population of speakers of languages other than English. However, it was observed that the use of Spanish localisation for this genre is still marginal is compared to multinational corporate websites in the US, where up to 74% of websites are localised (Jiménez-Crespo 2010b). Globally, less than 50% of non-profits have an English website, and within that percentage, 6.68% offered some content in Spanish, while only 2.54% offered a fully localised functional website.
The study provided a description of the genre in its source form, with eight conventional communicative blocks. It was observed that the main function of the website is twofold, mainly expository in order to provide information to potential users, and a secondary exhortative or persuasive function in order to persuade potential donors or volunteers and to provide a good public image of the organisation.
Regarding the differences between source and localised websites, the study clearly showed that localised ones show lower levels of interactivity and dynamicity than original sites. Among the blocks and sections with the highest rates of localisation loss were those related to dynamic content updates, such as ‘Press’ or ‘Calendar,’ as well as those related to interactivity, i.e. ‘Search’ or ‘Sitemap.’ The differences were also observed in the communicative situation in the description of the localised genre: some sections such as ‘Volunteer’ or ‘Foundation’ were among those with higher loss rates, and this is indicative of the different communicative situation in which the localised version clearly addresses a collective or group that might not necessarily be targeted for donations or volunteer work.
It should be mentioned that this study has merely focused on one aspect in which localisation loss might take place, the textual superstructure, and that the decisions related to these issues might be taken by commissioners or initiators, and not necessarily by translators themselves. In this sense, this descriptive study has to be understood within the wider recent sociological turn in Translation Studies that, as Michaela Wolf indicates:
[…] comprises the cluster of questions dealing not only with the networks of agents and agencies and the interplay of their power relations, but also the social discursive practices which mould the translation process and which decisively affect the strategies of a text to be translated (Wolf 2010: 29).
This is precisely the case in this study, since the main decisions regarding the textual configuration of localised websites might equally be shared among the many agents in the process. The final localised version is therefore the result of a complex interplay of commissioners, initiators, and translators that would require further study. It should be noted, therefore, that the notion of translation loss analysed in this study does not reflect conscious or subconscious decisions of individual translators, an aspect that could also be studied in the compiled corpus. Rather, this study highlights the degree of loss due to complex sociological issues in the localisation process. Many of these issues would require further study, such as the actual appearance of omission both as a translation strategy or as a translation error, issues related to quality and accessibility, etc. It is hoped that this paper sparks additional descriptive studies of those digital and printed genres that fulfill a social role within societies. Generally, most contrastive studies focus on those genres with the higher translation rates that are normally motivated by economic rather than social issues. The study of social translation or the social impact of localisation (Anastasiou and Schäler 2010) is not new in the discipline, but the crucial role of translation in multilingual communities with constant immigrant fluxes requires further study due to the impact translation has on their social fabric.
Acknowledgements
This research was possible thanks to a grant by the Rutgers Research Council. I gratefully acknowledge the help of our research assistants Soledad Chacón and Andrew Villada.
Bibliography
- Alamán, Teresa (2003). “The banner ad as a netvertising subgenre: an analysis of form and content.” Santiago Posteguillo et al. (eds) (2006). Internet in linguistics, translation and literary studies. Castellón de la Plana: Publicaciones de la Universidad Jaume I, 375-390.
- Anastasiou, Dimitra and Reinhard Schäler (2010). “Translating Vital Information: Localisation, Internationalisation, and Globalisation”. Journal Syn-theses, 13-27.
- Askehave, Inger and Anne E. Nielsen (2005). “Digital genres: a challenge to traditional genre theory.” Information Technology and People, 18 (2), 120-141.
- Baker, Mona (2004). “A corpus-based view of similarity and difference in translation.” International Journal of Corpus Linguistics, 9 (2), 167-193.
- — (1995). "Corpora in Translation Studies: An overview and some suggestions for future research." Target 7, 223-243.
- — (1993). "Corpus Linguistics and Translation Studies: Implications and Applications." Mona Baker, Francis Gill and Elena Tognini-Bonelli (eds) (1993). Text and Technology: In Honour of John Sinclair. Amsterdam/Philadelphia: John Benjamins, 233-250.
- Bassnet, Susan (2002). Translation Studies. New York/London: Routledge.
- Berkenkotter, Carol and Thomas Huckin (1995). Genre knowledge in disciplinary communications. Cognition, culture, power. New Jersey: Lawrence Erlbaum.
- Biber, Douglas (1993). “Representativeness in corpus design.” Literary and Linguistic Computing, 8 (4), 243-257.
- Bhatia, Vijay K. (1993). Analysing genre. Language use in professional settings. London: Longman.
- Bolaños Medina, Alicia (2003). “Dificultades y estrategias de la localización de sitios web comerciales del ingléss al español.” Santiago Posteguillo et al. (eds) (2003). Internet in linguistics, translation and literary studies. Castellón: Universitat Jaume I, 241-252.
- Bouffard, Paula and Phillip Caignon (2006). “Localization et variation linguistique. Vers une geolinguistique de l´espace virtuel francophone”. Meta 51 (4), 806-823.
- Brinck, Tom, Darren Gergle and Scott D. Wood (2003). Usability for the Web. San Francisco: Morgan Kauffman.
- Capelli, Gloria (2008) “The translation of tourism-related websites and localization: problems and perspectives.” Annalisa Baicchi (ed.) (2008). Voices on Translation, RILA Rassegna Italiana di Linguistica Applicata. Roma: Bulzoni Editore, 97-115.
- Classe, Olive (2000). Encyclopedia of Literary Translation into English. Chicago: Fitzroy Dearborn.
- De Beaudegrande, Robert and Wolfgang Dressler (1981). Introduction to Text Linguistics. London: Longman.
- Delisle, Jean (1993). La traduction raisonnée. Manuel d’initiation à la traduction professionnelle de l’anglais vers le français. Ottawa: Presses de l’Université d’Ottawa.
- Diéguez, Maria. I. (2008). “Análisis terminológico de sitios web localizados del inglés al español: uso de técnicas de amplificación y elisión.” Tradumática 6. ddd.uab.cat/pub/tradumatica/15787559n6a9.pdf (consulted 10.12.2011)
- Dunne, Keiran (2006). “A Copernican Revolution.” Keiran Dune (ed.) (2006). Perspectives on localization. Amsterdam and Philadelphia: John Benjamins, 1-11.
- Fernández Costales, A. (2008). “Translation 2.0. The localization of institutional websites under the scope of functionalist approaches.” Dries de Crom (ed.) (2009). Translation and the (Trans)formation of Identities: Selected Papers of the CETRA Research Seminar in Translation Studies 2008. http://www.kuleuven.be/cetra/papers/papers.html (consulted 10.12.2011).
- Fernández Sánchez, Franciso (2004). El folleto de cursos de idiomas para extranjeros: Análisis contrastivo (alemán-español) por tipos de emisor y subtextos. PhD dissertation. Universitat Pompeu Fabra.
- Gamero Pérez, Silvia (2001). La traducción de textos técnicos. Barcelona: Ariel.
- García Izquierdo, Isabel (2000). Análisis textual aplicado a la traducción. Valencia: Tirant lo Blanc.
- Göpferich, Suzanne (1995). Textsorten in Naturwissenschaften und Technik. Pragmatische Typologie- Kontrastierung-Translation. Tubinga: Gunter Narr.
- Hatim, Basil and Ian Mason (1990). Discourse and the Translator. London –New York: Longman.
- — (1997). The Translator as Communicator. London: Routledge.
- Hurtado Albir, Amparo (2001). Traducción y Traductología. Introducción a la traductología. Madrid: Cátedra.
- Janoschka, Anna (2003). Web Advertising. Amsterdam/Philadelphia: John Benjamins.
- Jeney, Cynthia (2007). Writing for the Web: a Practical Guide. Columbus, OH: Pearson Prentice Hall.
- Jiangbo, He and Tao Ying (2010). “Study of the translation errors in the light of the Skopostheorie. Samples from the websites of some tourist attractions in China.” Babel 56 (1), 35-46.
- Jiménez-Crespo, Miguel A. (2008). El proceso de localización web: estudio contrastivo de un corpus comparable del género sitio web corporativo. PhD Dissertation. University of Granada.
- — (2009a). “Conventions in localisation: a corpus study of original vs. translated web texts.” JoSTrans: The Journal of Specialised Translation 12, 79-102.
- — (2009b). “The effect of Translation Memory tools in translated web texts: evidence from a comparative product-based study." Linguistica Antverpiensia, 8, 213-232.
- — (2010a). “The Intersection of Localization and Translation: A Corpus Study of Spanish Original and Localized Web Forms.” Translation and Interpreting Studies, 5 (2), 186-207.
- — (2010b). “Web Internationalization strategies and translation quality: researching the case of ‘international’ Spanish.” Localization Focus—The International Journal of Localization, 8, 13-25.
- — (2011a). "To adapt or not to adapt in web localization: a contrastive genre-based study of original and localised legal sections in corporate websites." JoSTrans. The Journal of Specialised Translation, 15, 2-27.
- — (2011b). “From many one: novel approaches to translation quality in a social network era.” Linguistica Antverpiensia, 12, 153-174.
- Kennedy, Allistair and Michael Shepherd (2005). Automatic Identification of Home Pages on the Web. Proceeding from the XXXVIII Annual Hawaii International Conference on System Sciences, Big Island, Hawaii. Los Alamitos, CA: IEEE-Computer Society.
- Kussmal, Paul (1997). “Text-Type conventions and Translating: Some Methodological Issue.” Anna Trosborg (ed.) (1997). Text Typology and Translation. Amsterdam-Philadelphia: John Benjamins.
- Landow, George (1992). Hypertext: The convergence of contemporary Critical Theory and Technology. Baltimore: The John Hopkins University Press.
- Laviosa, Sara (2010). “Corpora.” Yves Gambier and Luc van Doorslaer (eds) (2010). Handbook of Translation Studies, Gambier, Yves and Luc van Doorslaer. Amsterdam-Philadelphia: John Benjamins, 80–86.
- Laviosa, Sara (1998). “The Corpus-based Approach: A New Paradigm in Translation Studies.” Meta 43 (4), 474–479.
- Lewis, David. K. (1969). Convention. A Philosophical Study. Cambridge, MA: Harvard University Press.
- Malmkjær, Kirsten (1998). “Love thy neighbour: will parallel corpora endear linguists to translators?.” Meta 43 (4): 534-541.
- Miller, Carolyn R. (1984). “Genre as social action.” Quarterly Journal of Speech 70: 151-67.
- Muñoz Martín, Ricardo (1995). Lingüística para traducir. Barcelona: Teide.
- Nielsen, Jacob and Loranger, Hoa (2006). Prioritizing Web Usability. Indianapolis: News Riders.
- Nielsen Jacob and Marie Tahir (2002). Homepage usability: 50 Websited deconstructed. Indianapolis: News Riders.
- Nord, Christiane (1991). "Scopos, loyalty and translational conventions." Target 3 (1): 91-109
- — (1997). Translating as a Purposeful Activity. Functionalist Approaches Explained. Manchester: St. Jerome.
- — (2011). “Making the source text grow. A plea against the idea of loss in translation.” Claudia Buffagni, Beatrice Garzelli, Serenella Zanotti (eds) (2011). The Translator As Author: New Perspectives on Literary Translation. Berlin: LIT Verlang, 21-29.
- Olohan, Maeve (2004). Introducing Corpora in Translation Studies. London: Routledge.
- Paltridge, Brian (1997). Genre, Frames and Writing in Research Settings. Amsterdam/ Philadelphia: John Benjamins.
- Pierini, Patrizia (2006). “Quality in Web Translation: An Investigation into UK and Italian Tourism Web Sites.” Jostrans, The Journal of Specialised Translation, 8, 85-103.
- Pym, Anthony (2010). Exploring Translation Theories. New York/London: Routledge.
- Renau Renau, Maria L. (2004). La página web de la industria cerámico-azulejera: un estudio descriptivo y su aplicación a la enseñanza del ingles para la informática. Doctoral Dissertation. Universidad Jaume I.
- Santini, Marina (2007). Automatic Identification of Genre in Webpages. Unpublished Doctoral Dissertation. University of Brighton.
- Shepherd, Michael and Carloyn Watters (1998). “The evolution of cybergenres.” Proceeding from the XXXI Hawaii International Conference on System Sciences, edited by Ralph Sprague. Los Alamitos, CA: IEEE-Computer Society, 97-109.
- Shepherd, Michael, Carolyn Watters and Alistair Kennedy (2004). “Cybergenre: Automatic Identification of Home Pages on the Web.” Journal of Web Engineering 3 (3-4), 236-251.
- Schmid-Isler, Salome (2000). “The language of digital genres—a semiotic investigation of style and iconology on the World Wide Web.” Proceedings of the 33rd Hawaii International Conference on System Sciences (HICSS 2000). http://csdl2.computer.org/comp/proceedings/hicss/2000/0493/03/04933012.pdf (consulted 08.12.2011).
- Shreve, Gregory M. (2006). “Corpus Enhancement and localization.” Keiran Dunne (ed.) (2006). Perspectives on Localization. Amsterdam-Philidelphia: John Benjamins, 309-331.
- Swales, John M. (2004). Research Genres: Explorations and Applications. Cambridge: Cambridge University Press.
- — (1990). Genre Analysis. English in Academic and Research Settings. Cambridge: Cambridge University Press.
- Vázquez-Ayora, Gerardo (1977). Introducción a la traductología, Washington: Georgetown University Press.
- Vinay, Jean P. and Jean L. Darbelnet [1958](1977). Stylistique comparée du français et de l’anglais. Paris: Didier.
- Wolf, Michaela (2010). Translation ‘going social’? Challenges to the (Ivory) Tower of Babel. MonTI 2, 29-46.
Websites
- European Parliament Proceedings Parallel Corpus 1996-2009. http://www.statmt.org/europarl/ (consulted 10.12.2011).
- JRC-Acquis of European legislation. http://optima.jrc.it/Acquis/ (consulted 10.12.2011).
- Open Subtitles corpus of movie subtitles.
http://opus.lingfil.uu.se/cwb/OpenSubtitles/frames-cqp.html (consulted 10.12.2011).
Biography
Miguel A. Jiménez-Crespo, PhD, directs the Master and BA program in Spanish Translation and Interpreting at Rutgers University, USA. He completed a PhD and MA-BA degree in Translation and Interpreting at the School of Translation and Interpreting, University of Granada, Spain. He has also studied at the University of Glasgow and Moscow State Linguistic University, where he taught Spanish and Translation. He has published numerous articles in international journals such as Target, Perspectives, Meta, Linguistica Antverpiensia and Jostrans. His research concentrate on localisation, corpus-based translation studies, translation quality, translation technology and translation training.
Note 1:
Translation from Spanish by author.
Return to this point in the text
Note 2:
Commission or ‘translation brief,’ according to functionalist theories of translation, refers in the instructions provided by the client (Nord 1997).
Return to this point in the text
Note 3:
New Jersey (27%) and New York (28%) both share a similar percentage of Hispanic population. Other states with highest population of Hispanics are New Mexico (37%), Texas (31%), Hawaii (27%), and Arizona (26%).
Return to this point in the text