Death to Neologisms : Domestication in the English Retranslations of Thomas Mann ’ s Der Tod in Venedig

Thomas Mann’s Der Tod in Venedig (1912) owes much of its fame in English to a translation from 1928 by Helen Tracy Lowe-Porter. The novella however has in fact been translated many times – first by Burke (1924, with a revised edition following in 1970), and, after Lowe-Porter, by Luke (1988), Koelb (1994), Appelbaum (1995), Neugroschel (1998), Chase (1999), Heim (2004), Doege (2007) and Hansen & Hansen (2012). Most of these versions are neither known to readers nor discussed in academic literature. This paper, which comes as part of a larger study on linguistic creativity in Der Tod in Venedig, focuses on the use of neologisms by Mann and what happens to them in (re)translation. Relying on a digital corpus composed of the complete set of English retranslations and a corpus-based methodology, the paper argues that, despite the extended time period between the publications and different translation conditions, neologisms are treated uniformly by the translators. Mann’s coinages are nearly always obliterated through normalisation and, if preserved, demonstrate less creativity overall than in the ST, raising questions about the Retranslation Hypothesis (RH) which proposes that early TT versions tend to domesticate while later ones increasingly foreignise.


Introduction
When Thomas Mann was writing Der Tod in Venedig (Death in Venice), he was, for various reasons, uncertain of its success and, after a first publication in two installments in the literary magazine Die Neue Rundschau (1912), initially only planned a limited print-run as part of Hans von Weber's Hyperionverlag Hundertdrucke.More than a century later, Der Tod in Venedig is one of Mann's most widely read works and has been translated into more than twenty-three languages.Its translation history in English is particularly interesting as it spans more than ten decades and eleven versions that include a revision, American and British editions and even a fan translation that was self-published online.The first translation, by Kenneth Burke, appeared in 1924 and was revised in 1970.It was soon followed by Helen Tracy Lowe-Porter's now infamous version in 1928, commissioned as part of an exclusive agreement between Fischer and Knopf, Mann's German and English publishers.Further versions came much later, a critical translation by David Luke in 1988 setting off a new surge of English Venices: Clayton Koelb (1994), Stanley Appelbaum (1995), Joachim

Retranslation and the retranslation hypothesis
As a study that includes all the English translations of a single work, this project falls under the phenomenon of retranslation, which describes "the act of translating a work that has previously been translated into the same language, or the result of such an act, i.e. the retranslated text itself" (Gürçağlar 2009:  233).While literary retranslation is common and, with certain text genres such as plays, even a prevalent phenomenon, research into the subject is relatively recent and still quite limited, especially when involving complete sets of target texts rather than only singular retranslations.Early research in the 1990s viewed literary retranslation as something positive, starting with Paul Bensimon (1990)  and Antoine Berman (1990) who both proposed, in separate pieces in the same volume of the journal Palimpsestes, specific ideas that were later referred to as the Retranslation Hypothesis (RH).Bensimon sees distinct differences between first and later retranslations, while Berman " (1990, as referenced by Gürçağlar  2009: 233) describes translating as "an 'incomplete' act" that "can only strive for completion through retranslation as each subsequent version not only increases the number of interpretations of the ST but gets closer to it.This idea of a "unidirectional move towards 'better' target texts" (Gürçağlar 2009: 233) was criticised a decade later, with multiple studies (Pym 1988; Chesterman 2000;  Koskinen and Paloposki 2003; Paloposki & Koskinen 2004; Susam-Sarajeva  2003 and 2006; Milton & Torres 2003; Brisset 2004; Brownlie 2006; Deane  2011; Deane-Cox 2014) arguing that retranslation is a much more complex phenomenon and that "historical context, norms, ideology, the translator's agency and intertextuality" (Gürçağlar 2009: 233) must all be considered.These studies challenge the traditional views about retranslation, including the assumption that early translations are always domesticating, while later ones increasingly foreignise; that translations always age; that the need for retranslation is directly driven by the passage of time and that a single factor can account for the decision to retranslate. 1 As this study focuses only on a single, specific feature (neologisms) in Mann's Der Tod in Venedig, it is not intended as a thorough examination of the Retranslation Hypothesis but offers data from one set of retranslations extending over almost ninety years that raises questions about the view that early and later versions differ in their translational approach in terms of domestication versus foreignisation.Similarly, although some of the discussed domestication strategies (in particular normalisation and explicitation) have been proposed as translation universals,2 the study does not position itself within this framework.Rather, it analyses the choices made by individual translators and within a specific case study only without considering whether these might possibly be typical and source language/target language-independent tendencies in translation.

Defining neologisms
The question that must be addressed first is What are neologisms?The answer is not straightforward.In simplest terms, neologisms are linguistic items that have been newly created by a language user.They have not been adopted into common language usage and occur infrequently.They may appear in a text only a single time, a phenomenon sometimes known as "hapax legomenon", but may also be used by individuals other than the item's original creator.Such utilisation is, by necessity, restricted in terms of quantity of usages and users employing the item in discourse.Neologisms contain an element of newness either in form, meaning or function.The first may be more prevalent and can manifest itself in manifold ways.It may involve making a word from scratch by assembling characters of a language in an unfamiliar manner but more often may rely on taking existing lexical items, either in part (morphemes) or whole (complete words), and combining them with other lexical components in a novel fashion to create a coinage.Newness however can also establish itself through meaning.In such neologisms an already existing lexical item is appropriated and given a completely new or an additional meaning distinct from any others attached to the word.Unlike neologisms by form, which language users are able to notice at a glance, these kinds of coinages may be overlooked easily as they are likely to establish themselves only over time, i.e. through an item's original meaning potentials transforming gradually or a new meaning developing through a word's usage in different contexts.Meanwhile, with neologisms in function, an existing item undergoes, either gradually or spontaneously, a grammatical change, transforming from one lexical category into another, for example, from a noun into a verb. 3The meaning of such coinages will normally be closely linked to the source item and, in most instances, the original and the new word will co-exist.They can however be easily distinguished on the basis of their immediate textual context (i.e. the clause or sentence they are used in) as well as word-class specific inflections, which will reveal their grammatical function.
The neologisms considered in this study are primarily those manifesting newness in form.Neologisms in meaning, meanwhile, were not considered suitable as they evolve in a different manner that makes them more challenging to identify immediately, certainly within a single source text and on the basis of a methodology relying in part on a wordlist (see Methodology Section 4).Meanwhile, items that feature newness in function can be identified more readily than neologisms in meaning, as inflections specific to a word's grammatical category are often involved in such cases.Textual context may however still be required, which an alphabetically sorted word list does not provide.Furthermore, there is the question whether newness in function is sufficient for an item to qualify as a neologism, given the much closer connection to existing words in comparison in particular to neologisms by form but arguably also neologisms by meaning.The issue that arises here is one of the degree of novelty and is linked to productivity, a feature explored in more detail in the next Section 3.1.2Inthis study newness in function was generally not deemed a sufficient criterion by itself for an item to qualify as a neologism, and such words were only included on occasion, i.e. when additional factors were present to justify the inclusion, such as a fixed phrase undergoing a grammatical change or an item exhibiting newness not only in function but also on another level.Such neologisms in functions were thus anomalies and judged as they occurred.
Processes of word formation are language-specific.Although many are common across languages, some ways of word formation are conventional only in particular languages.In a study concerned with linguistic creativity, this distinction between conventional and atypical word formation processes is important, as speakers may use the latter precisely to be creative.Conventional word formation processes however may also result in linguistic creativity.Grammatik der deutschen Sprache (Konitzer 1999) lists six maini.e.widely usedprocesses of word formation in German: Zusammensetzung (compounding), Zusammenbildung (synthetic compounding), Ableitung (derivation), Umbildung (conversion), Kürzung (abbreviation) and Terminologisierung (terminologisation), with compounding, derivation and conversion being of relevance here. 4 3 The creation of neologisms, whether in form, meaning or function, virtually always involves content words.Content words (also sometimes known as lexical words), which include adjectives, adverbs, nouns, verbs and interjections, are characterised as open word classes, allowing for additions.Function words, such as conjunctions, prepositions and pronouns, meanwhile belong to a closed word class, which contains fewer lexical items and rarely sees expansion. 4There is no space in this article to go into detail about the different word formation processes.
For a description and examples see the chapter on word formation in Konitzer's Grammatik der deutschen Sprache ("Die Wortbildung", 341-374).

Determining neologisms: Productivity, origin, level of innovation
While word formation processes create new words, these are not always neologisms as defined here.The decisive factor is whether the manner in which a word is formed is productive or not, and, if so, in what manner and degree.In the Hentschel grammar Deutsche Grammatik (2010: 282), productivity, which is language-specific, is defined as "ein Wortbildungsmuster, wenn es noch aktiv ist und neue Wörter danach gebildet werden können".Two words are key here: Muster (pattern) and aktiv (active).Productive word formation means that a pattern must be present.While 'pattern' can theoretically refer to top-level word formation categories (i.e.compounding, conversion, et cetera), it can also describe a specific recurring process within these larger groupings, e.g.adjective-to-noun derivation through the addition of the suffix -heit (schön -Schönheit, dunkel -Dunkelheit).A pattern by itself, however, is not sufficient to speak of productivity: the pattern must also be in use, that is, it must be active.Productivity is not a permanent quality: word formation patterns can, over time, fall out of usage, something which language users who wish to be creative may take advantage of.
Active is also a relative term, as some word formation patterns are wide-spread and may, particularly in spoken language, lead to new words being coined impromptu every single day, while others may be comparatively rare.When word formation patterns are shared across languages, they are often more productive in one language than in another, meaning that items formed by the same process may go unnoticed in one instance but will be considered creative coinages in the other.A good example here is the practice of conversion, which is used both in German and English (as well as other languages) but is highly productive in German, in particular in the form of nominalisation, with the pattern being common in all kinds of linguistic contexts, from oral speech to newspaper articles to poetry.Words formed by conversion in German are thus often not considered creative, while in English they are perhaps more likely to be so.
Productivity is not an absolute factor.It constitutes a useful criterion in the consideration of creativity and for determining whether something is a neologism or not.The absence of productivity, however, does not necessarily signify creativity since a word, at least theoretically speaking, may have a completely unique formation but may have been adopted into common language usage.Similarly, even the most productive processes do not mean an automatic exclusion, as exceptions are always possible, for example, when a fixed phrase undergoes a conversion, or, as happens frequently, when a neologism is the result of not one but several word formation processes in combination.In such instances native speaker intuition is crucial in deciding whether an item qualifies as a creative coinage or not.What matters thus is not just the degree of productivity present, but the exact context for each item concerned.
Furthermore, it may be helpful to consider a word's origin and history of usage, as these may give insight into whether it is a neologism or not.Etymological and diachronic research is not always straightforward.While the term neologism refers to something that has been newly coined and thus suggests a specific creator, it is more often than not difficult to attribute a coinage to the individual that first made and used it, as well as the context in which it first appeared, in written but particularly so in oral discourse, which for a long time did not leave any record.It may be possible to approximate the time period in which a lexical item first appeared, but determining the exact moment of genesis and the historical trajectory of a word would require detailed study without any guarantee that a word's origin will eventually be established. 5Moreover, definite attribution is complicated by the fact that neologisms may have been used by different individuals in different instances at roughly the same time.However, as long as we accept that, with the exception of impromptu coinages, a neologism is not necessarily a word that is one hundred percent new, nor created and used by only one and the same single individual, this need not be disconcerting.Neologisms may indeed have some spread, although the frequency and range of usage have to be limited as they cannot be words that are clearly on the way of being adopted into a language.6They must also be carefully distinguished from items that are used infrequently for other reasonse.g.specialist terms from specific fields, advanced level synonyms of core or general usage vocabulary or old-fashioned items that are slowly recedingas these, unlike neologisms, have been adopted into the language, even if they are not used every day or not known by most speakers.A word's status, whether as a neologism or as a lexicalised item, is therefore not permanent: coinages may spread and become part of the general language over time, or they may only be used a handful of times by their creator, only to then disappear completely.Equally, long-established and once popular words can fall out of usage, something that is important to remember in a study involving a source text published more than one century ago and translations spanning as far back as 1925.
Finally, it needs to be noted that words differ in their individual level of innovation: although we may classify two items as neologisms, one may be more novel than the other, e.g.Lewis Carroll's nonce words brillig, toves, outgrabe are more striking coinages than a relatively self-explanatory compound like Mann's Einzelinspiration. 7This applies even if their creation follows the same general word formation pattern or when items may seem near-identical on the surface.
In order to classify a lexical item as a neologism, thus multiple criteria are applied: the type of newness (form, meaning, function, or a combination of these) which the word exhibits has to be established, the word's productivity in its formation and its individual level of innovation have to be considered, as well neologisms like Einzelinspiration, which other users may coin without ever having read Mann's novella, they are unlikely to reoccur.
as its origin (if identifiable) and its (relative) spread and frequency of usage, all of which must be judged on a case-by-case basis.

Tod in Venedig Corpus (TIVC): ST versions and TTs
The study uses an entirely digital translational corpus, containing two subcorpora: a ST corpus with two texts and a TT corpus with eleven texts.The former includes two of three different versions of Der Tod in Venedig, namely the so-called Hundertdruck (HD, published by Hyperionverlag München in 1913) and the Buchausgabe (BA, published by S. Fischer also in 1913),8 while the latter consists of all currently available translations, including Burke's original and revised versions.Both the BA and the HD are part of the corpus, as one translation (Doege's) uses the latter as its source.The specific TT editions are given in the bibliography. 9In most instances paper copiesoften obtained second-hand as many translations are no longer in printwere used, meaning that digitisation (through scanning) and conversion into machine-readable files with Optical Character Recognition (OCR) software was necessary.Basic tagging was added to files for paragraphs, orthographic markings (italicisation) and foreign language words as well as metadata including (as applicable) text name, author, translator, year of translation, language and/or regional variety (German, British or American English), and source text.Any paratextual materialusually footnotes and, in the case of e-book versions, also hyperlinked annotationswas removed.
The corpus itself exists in two main forms: as WST-TIVC, a digital corpus created with Mike Scott's widely used linguistics software suite Word Smith Tools, and as A-TIVC, an aligned corpus in MS Word.10

Methodology for determining neologisms
Neologisms were determined through a three-fold method: to begin with, intuitive judgment was used to draw up an initial list (L1) of potential coinages, which were then subjected to a first, more objective cross-check through corpus resources, with remaining items (L2) being checked more extensively a second time with corpus resources, resulting in a final list (L3).

Intuitive judgment (L1)
The starting point for the initial list of neologisms was a ST wordlist drawn up with WST tools.On the basis of native speaker intuition (i.e. the author's) neologisms were identified and placed on a tentative list (L1), with a general approach of inclusiveness guiding this pre-selection, i.e. all items that appeared like potential coinages were included.This original list contained 253 items, which required further confirmation both due to intuition being a subjective and thus potentially problematic criterion for identification as well as the expectation that the all-inclusive approach would result in at least some incorrectly identified items.

Corpus-based cross-check (L2 and L3)
Further confirmation was done through a double cross-check of each L1 item through several independent corpus resources of different types and compositions.This second and third check11 provided an alternative to the essentially subjective use of intuition of a single individual (i.e. the researcher), although it has to be noted that corpus resources are not entirely free of subjectivity either: they are merely more objective.
The range of resources included 1) dictionaries, both of the monolingual and bilingual kind, 2) an encyclopaedia, 3) general corpora and 4) web search engines, containing both contemporary but in some cases also historical reference data.A detailed description follows in Section 4.2.3.The resource types were included for distinct reasons, with each type providing an indication about the status and usage (or lack thereof) of words in a language.With words being created and adopted into a language in different ways, the use of a combination of resource types was considered most suitable for determining which list items were neologisms and which not.The cross-check of items was resource-specific, as the resources serve distinct purposes.Dictionaries, for example, will generally have a single entry for each item and provide information on its meaning potential(s) and, usually, a few usage examples.Additionally, synonyms and antonyms may be given.Corpora, meanwhile, contain collections of spoken or written texts that are extensive both in overall quantity and individual size, so that usage patterns of queried items may be revealed.These distinctive purposes mean that with some resources (i.e.dictionaries and encyclopaedia) it is primarily the occurrence of an item that matters when determining an item's neologism status, while with others (corpora, web search engines) the quantity of occurrences is most important.However, as all resources used are digital, quantity of occurrence in fact constituted an additional measurement applicable for the dictionaries and the encyclopaedia consulted, as, unlike with traditional print resources, not only entries but also usages are retrieved.To illustrate: with many electronic resources, including all the ones used in this study, a query for a particular lexical item (e.g.'house') will retrieve hits which are entries, with entries being the specific, individual listing for the item, but also its usages, i.e. all the instances the item appears within entries.Depending on the makeup of the specific resource, usages may furthermore include appearances within the acknowledgments, preface, appendix, et cetera. 12re details on the significance of occurrence versus quantity of occurrences are provided in the overview of cross-check resources.

Dictionaries
Four dictionaries were consulted for the cross-check: Duden, OWID, Pons and Leo, with the first two being a monolingual and the latter two bilingual (English-German) resources. 13All these dictionaries are large in size and include keywords from a wide range of texts.The Duden, which has been published since 1880, is the leading dictionary for the German language.Its most recent print edition (2013) contains 135,000 keywords while the online version (<http://www.duden.de>)has more than two million word forms.14OWID (Online-Wortschatz-Informationssystem Deutsch, <http://www.owid.de>) is managed by the German language research institute IDS (Institut für Deutsche Sprache) and contains 300,000 keywords in its main component, elexico. 15ons (<http://www.pons.de>),maintained by a German publisher of the same name that is specialised in dictionaries and language learning materials, amounts to at least 120,000 keywords.Leo (<http://dict.leo.org>), which was originally established in 1995 by members of the computer science department of the Technische Universität München as a resource for German-English translation and is now owned by an independent company (Leo GmbH), stood at 795,020 keywords (December 2014), with new ones being added on a continuous basis.Although the Duden and Pons also exist in print form, the web-based resources were used in all instances.
Traditionally, dictionaries list lexical items in the base form specific to each item's word category (e.g.verbs as infinitives, nouns in singular and, for German, in nominative case) and are then sorted alphabetically.With the primary function of dictionaries being to serve as a reference guide for the meaning potential of lexical items, each item is normally listed only once. 16As already indicated, this characteristic also means that what matters in terms of neologisms, is whether a word appears in a dictionary, not the quantity of occurrencesin contrast to some of the other resources used for the crosscheck.Any item that is listed in a dictionary is unlikely to be a new coinage, as inclusion normally indicates that some level of lexicalisation or even conventionalisation of the term has taken place, i.e. that the item has been adopted into wider language usage beyond that of its creator or original text.This characteristic applies for both entries as well as usages, which provide further clues about a word's usage (i.e.whether it is a core word of the language, a specialised term or an obsolete item).Items that have both entries and usages, particularly if these are high in number, are clearly not neologisms indeed, exceptions to this observation are highly unlikely.
If occurrence corroborates that a lexical item has been adopted into the language more permanently, then non-occurrence must be indicative of neologism.While this non-occurrence is indeed an important criterion, it however does not automatically make a lexical item a neologism.No dictionary contains all the words of a language, for various reasons: there are restrictions in size (the number of items that can be included, particularly in print editions, is limited) but also in content, as dictionaries may have a specialised focus or may exclude items that are specific to a region (British vs.American English), a field (medicine, mathematics, literature), a usage group (youth jargon) or whether a word is primarily used in written or spoken contexts.Language is also dynamic and evolves continually.While some lexical items become obsolete over time and are eventually removed from newer editions, new lexical items will also be created, and may undergo lexicalisation until, if successful, they are added to a language's vocabulary.This process that may last anything from a few months or years (usually for content words) to decades or even centuries (function words).In other words, non-occurrence merely indicates a lexical item's potential of being a neologism.
In the cross-check the factor of non-occurrence due to omission guided the choice of dictionaries and motivated the decision to use several rather than one dictionary as well as to consult dictionaries with a large number of keywords.With the ST not belonging to a field with a specialised terminology, general dictionaries containing words from a wide context were deemed more appropriate than specialised ones.Additional factors determining the selection were online availability and free access to each dictionary used.

Encyclopaedias
Encyclopaedias are not typically intended as a linguistic resource but provide factual information about a wide range of subjects rather than ordinary lexical items.While there is likely to be some overlap between encyclopaedias and dictionaries in the items included, with the former specialist terms are more prevalent.However, these conventions are changing with online encyclopaedias.With virtually no limitations in size and the possibility to expand and update entries continually, web-based encyclopaedias have significantly shifted in content focus, featuring anything and everything from current events to celebrity profiles.The encyclopaedia consulted for this study also contains entries which are normally restricted to dictionaries, such as for personal pronouns or articles, narrowing the distinction between these different kinds of resource types.Additionally, it features items that are normally found neither in general encyclopaedias nor dictionaries, such as 0 (the integer) or symbols (∅,  disambiguation).This content shift is insignificant for this study.Indeed, it may in fact increase the suitability of the resource type as the data for the cross-check thus becomes more diverse.Additionally, as with online dictionaries, digital enyclopaedias share features with electronic corpora that print editions do not.
The encyclopaedia used in the neologism cross-check for this project was Wikipedia, both because of its immense (and continually increasing) number of entries as well as the fact that its query function retrieves not only entries but any occurrence of the queried item within each entry.Wikipedia is exclusively web-based.It was first launched in 2001 and is now available in 286 different dialects and languages.Its German subcomponent (<http://de.wikipedia.org/>),which by default is searchable separately from other Wikipedia sections, currently has more than 1.8 million entries (March 2015). 17The total number of tokens composing these entries is not known.
For the cross-check, it is both occurrence and quantity of occurrence that matter.The former is most significant (and, indeed, in print encyclopaedia, it is the only measurement available) as any item featured in an encyclopaedia will have some level of conventionalisation.That is, entries are included precisely because they are in use, even if only rarely or exclusively within a specific field, so that individuals not familiar with them may look them up.In other words, any item that has an encyclopaedic entry is highly unlikely to be a neologism.With the digital Wikipedia it is possible to apply quantity of occurrences as a second measurement that can further corroborate a word's status, and, in cases where there are usages but no entries, this measurement may be decisive. 18

General corpora
Corpora were the only resources used in the cross-check that were specifically designed for linguistic research.As with dictionaries and encyclopaedias, the choice was for large corpora containing as many lexical items from general 17 At the time the data was retrieved, it stood at more than 1.5 million entries (May 2013). 18To obtain quantities, Wikipedia users must query the database either by following a particular path from the left-side column (Werkzeuge → Spezialseiten → Suche for the German Wikipedia section) or by clicking the magnifying glass symbol in the Search-box on the top right of the page, as direct queries via the search boxes either on the front portal for all of Wikipedia (<www.wikipedia.org>) or on the language-specific front pages currently do not provide any numbers (May 2013).language usage as possible.Two general, monolingual corpora of the German language were selected: the DWDS and COSMAS II.The DWDS (short for Digitales Wörterbuch der deutschen Sprache, <http://www.dwds.de>) is an online resource owned by the Berlin-Brandenburgische Akademie der Wissenschaften.The DWDS is, as its name already indicates, not only a corpus in the strictest sense of the word but includes a general dictionary, an etymological dictionary, a thesaurus and a word profiler application in addition to a large corpus consisting of three different sections (Referenzkorpus, Zeitungskorpus and Spezialkorpus) with multiple subcorpora each.Most texts date from the twentieth century.In the cross-check the primary focus of evaluation was on data from the Kernkorpus 20/21 (123 million tokens) and Zeit & Zeit Online (226 million tokens) subcorpora.
COSMAS II (Corpus Search, Management and Analysis System, <http://www.ids-mannheim.de/cosmas2/web-app/>) is a project of the Institut für Deutsche Sprache (IDS), which is available as an application for different platforms.In this case the web-based resource was used.Its data come predominantly from the Deutsches Referenzkorpus (DeReKo, <http://www.sfs.uni-tuebingen.de/dereko/>),a government-funded joint research project of several academic institutions, as well as from IDS project groups.The corpus currently contains 7.3 billion tokens within 108 subcorpora and includes a wide range of texts: newspapers, airline magazines, bibliographical literature, twentieth and twenty-first century prose fiction, historical texts and encyclopaedia entries.
With this resource type, occurrence but especially quantity of occurrence was important when evaluating an item's neologism status.Unlike with the previous two resources (dictionaries and encyclopaedias), the purpose of corpora is not to provide single entries with definitions and explanations but to retrieve all examples of a lexical item from its database of texts to provide insight into its language usage patterns.As general corpora contain texts of many different types and from a wide range of sources to be as representative of language in general as possible, a single occurrence of a lexical item does not automatically disqualify it from being a neologism nor makes it 'highly unlikely' for it to be a coinage, as applied for dictionaries and encyclopaedias.A single or even a few individual hits may well be examples of a neologism in use, for example, in a literary text or in a newspaper article, both texts types which in German will often feature newly coined words.Each occurrence therefore needs to be checked manually, with the total quantity also becoming a determining factor: the greater the number of hits and the wider the range of texts in which the hits occur, the less likely it is that the item in question is a neologism. 19e combined use of the DWDS and COSMAS provided a sufficiently large database for the cross-check and the somewhat different focus in their 19 In the case of DWDS, although the primary focus was on data from Kernkorpus 20/21 and Zeit & Zeit Online, the cross-check also took into account whether a query retrieved hits in other subsections (the dictionaries, the thesaurus or the word profilers), as such occurrences were likely to rule out an item's neologism status.
components extended the range.The fact that both included historical subcorpora, with data from the time of publication of the ST and also before, was an added advantage, even more so since the dictionary and encyclopaedia sources were contemporary.Occurrence in both was not deemed necessary to declare a queried item as a non-coinage because quantity and range of occurrence within a single corpus were considered more important.Similarly, even if hits were retrieved from both the DWDS and COSMAS, this did not immediately disqualify an item as a neologism, even if in most instances it provided stronger evidence.Despite the immense total size of the two corpora together, non-occurrence was not enough to declare an item's neologism status; in this case confirmation from the other resource types was essential.

Web search engines
The final resource for the cross-check were web-based search engines.The web as corpusi.e.trawling the internet with the help of web search enginesis still a controversial tool in linguistic research.Although it is not the only crosscheck resource that is not primarily intended for linguistic analysis, it is likely the most problematic.While it is standard for dictionaries and encyclopaedias to include at least some information about their composition, the compilation process and the contributors involved as well as metadata about each document contained, there are no such conventions for online resources.The most widely used search engines are owned by multinational corporations, which closely guard details on how their tools operate, often resulting in a lack of transparency in terms of how search engines index, retrieve and sort data.The exact total size and content of indexed materials is unknown and, given the immensity and the ever-changing form of the internet, not just difficult but essentially impossible to ascertain.The constant and comparatively rapid change of online content also means that exact replication of queries is not an option due to the impermanence of the data retrieved and factors such as cookies and individual browsing history which influence web searches. 20rchives, such as the Wayback Machine (<www.archives.org>)for preserved webpages from the past, are incomplete and not directly linked to specific search engines.Additionally, any changes in a search engine's mechanisms (e.g. in the algorithms that sort results) may be less well documented or not made available to the public, in contrast with resources specifically intended for linguistic research. 21While web search engines are thus at best an imperfect resource, they are not completely useless and, at least in combination with other resources as done here, provide information that can assist in determining a lexical item's neologism status.
Three search engines were used in the cross-check -Google Search, Bing, and Yahoo!Searchas they are among the websites used most frequently (in terms of numbers of queries and users) as well as most widely (geographical 20 Given its medium, the data retrieved through web search engine changes more rapidly than print sources but probably also more than most other online resources. 21Indeed, the query language originally used to retrieve exact phrases (January 2013) seems to have changed (December 2014).reach). 22Google Search, owned by Google Inc., indexes billions of webpages which the search engine then trawls to retrieve keyword hits and list them in abbreviated form with links to respective source pages.Search results are sorted with the help of a patented algorithm-based mechanism called PageRank, although other factors may influence the ranking.Bing and Yahoo!Search, owned by Microsoft and Yahoo! Inc. respectively, proceed similarly. 23he percentage of total webpages indexed by any of these search engines is unknown and, indeed, probably unknowable.Non-indexable data (the so-called Deep Web), such as websites that require registration and login for access, unlinked content and non-html content, is normally excluded from search engine results.For all three search engines, country-specific sites and advanced setting to optimise results are available.However, unlike with all other cross-check resources used, none allows for lemma-based queries, meaning that at this stage of the cross-check base forms of the specific lexical items were used: infinitives in the case for verbs, singular forms in the nominative case for nouns and adjectives.The query language was specific for each search engine but consistently applied for each search with the particular search engine to maximise the comparability of the data.As stated previously, with this resource, quantity of occurrence was the most important factor.

ST neologisms: Classifications
The final list contained 107 neologisms that were classified into a number of distinct categories: there were compound coinages (either with or without a hyphen), neologisms by derivation, creative variants of existing forms as well as new creations that used a combination of strategies or were otherwise difficult to classify.An explanation of the different types of coinages used by Mann follows below.24

Compound coinages
Compounding, a common process of word formation in German, was present in the ST in two forms: as hyphenated and non-hyphenated compounds.Amtlich-Erzieherische (das); Einsam-Stumme (der); Einsam-Wache (der); Geschliffen-Herkömmliche (das); Göttlich-Nichtssagende (das); Mustergültig-Feststehende (das); Nebelhaft-Grenzenlose (das); Tapfer-Sittliche (das); Verheißungsvoll-Ungeheure (das)  1).Mann's hyphenated compounds, however, are not all equal and come in different forms, some of which are more creative than others.They may be mere spelling variants of existing words, as is easily revealed by removing the hyphen and joining the compound's components without any space in between: Bäder-Hotel  Bäderhotel, Hotel-Angestellte  Hotelangestellte and Granatapfel-Getränk  Granatapfelgetränk are some examples.Although these types of compounds are marked in the sense that they use non-standard spelling and were included in the final count, they arguably demonstrate stylistic preference more so than creativity.Other hyphenated compounds generally fell into two subcategories.A significant number (nineteen) were composed of two adjectives connected through the hyphen (ängstlich-übermütig, dumpf-süß, heilig-schattig, et cetera), the components being notably distinct in terms of meaning, to the point of being a seemingly nonsensical combination.The contrast emphasised the creativity on the part of Mann, with the unconventional compounds taking on both a new form and meaning.The other type of hyphenated compound, with nine examples present, was similar, consisting also of two adjectives but additionally involved nominalisation, thus resulting in forms such as (das) Nebelhaft-Grenzenlose, (das) Tapfer-Sittliche and (das) Verheißungsvoll-Ungeheure. Non-hyphenated compounds, i.e. the joining of two or more lexical items with no punctuation mark or space to delineate the 'border' between components were, with forty-three instances, used as frequently as hyphenated compounds (see Table 2).They virtually always were formed by two parts and were predominantly nouns (Allerweltsferienplatz, Bettelvirtuose), although a few adjective compounds were used also (dickdunstig, traumglücklich).Like with the hyphenated compounds some combinations were unconventional and striking, others were rather classic formations (e.g.Gepäckbeförderungsamt). Nonhyphenated compounds were different from hyphenated ones not only because they lacked hyphens.Compounds demonstrated greater diversity in their composition, joining nouns with adjectives (such as the traumglücklich mentioned above) or drawing on participles (schwergeschmückt).Notably, there were no nominalised adjective compounds.Unsurprisingly, the various types of compounds used by Mann exhibit different degrees of creativity.While it cannot be exactly quantified, it is noticeable.

Derivation beutelschneiderisch; Halbschurke (der); Tagedieberei (die) Table 3: Neologisms by derivation
Neologisms formed through derivation (Table 3), specifically affixation, constitute another type of neologism used by Mann.Affixationi.e. the adding of a morpheme to the stem of a word, either by placing a semantic unit before (prefix), within (infix) or after (suffix) of the itemis a somewhat questionable tool for creativity as the practice is often highly productive.Many affixes are used on a regular basis and have become conventionalised.Neologisms by affixation were thus not included outright.
While a handful appeared on L2 (Halbschurke, Tagedieberei,  beutelschneiderisch, grundsonderbar, übermodisch, korridorartig,  nonnenähnlich), only the first three made onto the final list.None of the items made an appearance in any of the first three cross-check resource types, with only web search engines retrieving data.The number of hits were comparatively similar for beutelschneiderisch, grundsonderbar, Halbschurke and nonnenähnlich (for Google 651, 358, 545 and 484 respectively) but somewhat to significantly higher for korridorartig, Tagedieberei and übermodisch (for Google 2503, 1690 and 931 respectively), providing good reason to exclude korridorartig and übermodisch on this basis.Although numbers were also high for Tagediebereiindeed, higher than for übermodisch -, the item made onto the final neologism list.Both in the case of Tagedieberei and also beutelschneiderisch, the root word was more unusual and affixation resulted in a more striking item.The Google hits also proved interesting, especially for Tagedieberei, where quite a few linked to empty dictionary hits, 25 to Der Tod in Venedig related textual materials (including links related to a 2012 Thomas Mann exhibition titled "Stegreifleben und Tagedieberei") as well as some websites where the word was used together with other obvious neologisms. 26As for Halbschurke, the data retrieved linked notably often to online versions of the novella, again making some hits irrelevant.With Halbschurke there was also a semantic difference that provided reason to include it on the final list: The prefixes of the omitted items grund-and über-are used for emphasis, while the suffixes -artig and -ähnlich indicate comparison.None of these, however, modifies the meaning of the word stem in a more profound way, something which does not apply for Halb-in this instance.It is a prefix that indicates quantity, but for a word that is not normally quantified in such manner: Schurke (rascal, wretch) refers to a person, yet a person cannot literally be halved, only metaphorically.The use of affixation is thus more novel in this manner and, with 25 Search engine queries included noise in the data retrieved.One type of irrelevant hit were empty dictionary hits, which, on Google's list, appeared to suggest a dictionary entry for the word queried but, once accessing the link, only revealed a query page with an error message indicating that the item was not found in the dictionary.hits being sufficiently low, justifies inclusion for this reason.While such an argument cannot be made for the other suffixes -isch (denoting associated qualities) and -ei (signalling an activity or the domain of an activity), and while at least with Tagedieberei hits were comparatively high, intuitive judgment ultimately overruled these concerns, with both Tagedieberei and beutelschneiderisch being classified as coinages.

Conversion
Befallene (der); Enthusiasmierte (der)*; gluthauchend*; keimbekämpfend; Rosenstreuen (das); Weitherkommende* (der) Conversion, as noted previously, is a particularly productive process in the German language and not necessarily creative.Der Tod in Venedig, unsurprisingly, contains many examples of conversions throughout (e.g.das Gesetzmäßige from the adjective gesetzmäßig, das Klappern and das Ausschlafen from the verbs klappern and ausschlafen respectively), but these were generally not included in the final list of neologisms due to productivity.Items that did make it onto the final list were all somewhat atypical, involving in place of a single word item, a short phrase (e.g.Keime bekämpfen  keimbekämpfend) or applied the conversion process to a word more than once (marked with an asterisk Table 4), as with Enthusiasmierter, derived from the word enthusiasmieren, which is turned, in the form of a past participle, into an adjective (enthusiasmiert) and then nominalised.Such formations arguably have a lower level of innovation than some other types of neologisms, but are nonetheless striking to users.

Creative variation on an existing form
Halbdame Table 5: Neologism as a creative variant of an existing form One item, Halbdame, demonstrated creativity by taking an existing lexical item but altering it in some way, something that we often also refer to as wordplay.In wordplay original items are generally still recognisable, although only to individuals familiar with them, with a change occurring both on the formal and the semantic level.While Halbdame on the surface looked like a neologism by prefixation (similar to Halbschurke above), it is in fact derived from the word Halbweltdame and, by extension, Halbwelt.The latter is defined as "eine elegant auftretende, aber zwielichte, anrüchige Gesellschaftsschicht" 27 in the online Duden, with a Halbweltdame being "eine der Halbwelt angehörende Frau" 28 .Mann's Halbdame appears in a sentence in Chapter 3, "Die 27 "A social class that appears elegant but is shady and infamous" (my translation).
28 "A woman that is a member of the Halbwelt" (my translation and emphasis).
Gouvernante, eine kleine und korpulente Halbdame mit rotem Gesicht, gab endlich das Zeichen, sich zu erheben" 29 (my emphasis), playing with the source words both in form and meaning.While Halbwelt and Halbweltdame connote the seedy underworld and are heavily negative, Mann's creation shifts meanings and connotations.The immediate sentence context contains some negativity in terms of the woman's physical appearance (she is "korpulent" and has a "rotes Gesicht") and also references social class, however, there is no indication, either in the text that frames the word immediately nor in the novella as a whole, that she belongs to a "zwielichte, anrüchige Gesellschaftsschicht".Although the suggestion is that she is of a different and lower social class than her employers, the negativity is somewhat lessened through the disassociation.Again it is notable that Halbdame, like Halbschurke mentioned earlier, cannot be interpreted literally, only metaphorically, making the coinage more striking on this level.
Table 6: Neologisms by combining multiple word formation processes Finally, a handful of items were more challenging to classify.Instead of relying on a single strategy for innovation, they make use of several simultaneously.Although there are not many such coinages in Mann's Venedig (see Table 6), this strategy of creative combination where multiple word formation processes are applied to a single item one after the other, was to be expected, as it reflects the organic development of language.In most instances the word creation was fairly straightforward, e.g.Stegreifdasein is formed from the fixed phrase aus dem Stegreif by first deconstructing the phrase into its parts and then combining the noun component with Dasein.With halbgeflüstert however, the exact order of the different formation strategies applied was not so clearly identifiable: the word might have been formed either by turning the verb flüstern into an adjective through its past participle form, geflüstert, with the prefix being added after.Alternatively, Mann may have taken the noun Geflüster as the starting point.
The different types of neologisms present in Der Tod in Venedig reveal the linguistic preferences of Mann, demonstrating not only a varied range of linguistic creativity but providing different kinds of challenges for the translators.
29 "The governess, a small and corpulent Halbdame with a red face, finally gave the sign to rise" (my translation and emphasis).

Creativity in the translation of neologisms
Data analysis reveals that Mann's neologisms are nearly always removed in the English versions of Der Tod in Venedig as all the translators demonstrate a strong preference for fluent translations. 30More than nine times out of ten coinages are replaced by lexically conventional options, while creative choices are the exception and occur not only infrequently but seemingly also arbitrarily.
As can be noted in Table 7 below, on average only 7.05 % of the neologisms on L3 are rendered creatively in translation, with Koelb most often using nonconventional solutions (in nine out of 107 instances, or 8.41 % of the time) and Neugroschel and Doege the least (in six instances, or 5.61 % of the time).The overall percentage average is not only strikingly small, but with less than 3 % difference between the most and least creative translations, the evidence is not sufficiently compelling to declare any one TT as 'more' or 'less creative' in terms of translating neologisms.What can be said, however, is that preserving neologisms is clearly not a priority in any of the target texts.Some variations can be seen with the type of neologism.Compounds, both with and without hyphens, are the most frequently occurring kind of neologisms.In translation, non-hyphenated compounds are rendered creatively more often (6.77%) than hyphenated ones (2.75 %), although both the percentages as well as the percentage difference between the two compound categories remain small.Within the hyphenated compound category, there is minimal variation among compounds that are spelling alternatives (SPA), double adjectives (DADJ) or nominalisations (NOM).The last are normalised in all instances, while spelling alternatives and double adjectives have preservation rates of 3.64 % and 3.35 % respectively (see also Table 8).The complete normalisation of nominalisation is somewhat surprising, given that in terms of degree of creativity in the ST these are more striking than spelling alternatives, whose inclusion on the neologism list is debatable, yet which at least remain creative in a handful of cases in the TTs.With other types of ST neologisms, the majority are also eliminated across all targets texts, but preservation is generally higher than with compounds: conversions are rendered creatively 13.64 % of the time, while rates are 18.18 % for combinations and 20.45 % for derivations.The only exception is the creative variant category, which is never translated innovatively.
The lower number of occurrence of all these types of neologisms compared to both compound types also makes it difficult to argue for any definite trends here.Overall, the decision of when to maintain creativity in the TTs seems to be largely random.Forms that are more creative are not necessarily more likely to be preserved (compare the data for different types of hyphenated compounds) and the distribution of TT neologisms is sporadic: in thirteen instances it is a single translation opts for a neologism, in four it is two translations, and in two instances it is three translations.Additionally, there are five examples 31 One neologism in the ST has, with eleven TTs, eleven potential instances where it could be rendered either creatively or normalised in translation.Hence, '32 out of 473' indicates 32 instances of creative translations in 473 potential moments, meaning that 32 creative forms appear across all TTs for 43 neologisms in the ST.(gluthauchend, breitschattend, halbgeflüstert, rotbewimpert, sargschwarz) with which nearly all translators use an innovative form, yet show interestingly little variation ( Burke's red-lidded for rotbewimpert is likely a mistranslation.Chase's deviation is more interesting as he is the only one who, in three instances, opts not only for a creative form, but for one that is not identical to those chosen by the majority of other translators.The repeated neologisms, meanwhile, raise the question of whether later translators may be copying earlier ones here.Although it can only be speculated whether the same coinages being reused indicates not creativity but a lack thereof or whether the repetition occurs for other reasons, 32 what is certain is that, in general, creativity is not only not a priority in the English versions of Der Tod in Venedig, but that the form it takes in translations is limited.Mann's neologisms fall into a number of different categories, whereas creative TT forms are, without exception, hyphenated compound words.Like in the ST, some of these are more striking than others, with some (e.g.travelingpad, gondola-landing) barely amounting to more than spelling alternatives and others occasionally taking more daring forms (coffin-black-varnished, cloudswollen, melancholy-enthusiastic), but no translator experiments with any other technique.

Normalisation in translation
The most common approach to neologisms in the TT corpus is thus normalisation, a strategy that Sara Laviosa describes as "the translator's sometimes conscious, sometimes unconscious rendering of idiosyncratic text features in such a way to make them conform to the typical textual characteristics of the target language" (2002: 54-55), with Mona Baker adding 32 As the examples in this chapter concern single lexical units rather than more extended phrases, limited variation may be explained, at least in some cases, by this factor.Rendering a word like 'sargschwarz' as 'coffin-black' is literal as well as sensible, and thus quite probable.
that it "exaggerate[s]" (qtd. in Laviosa 2002: 69) the typical target text features.Normalisation occurs either due to "systemic constraints of the target language" or "the translator's own preferences" (Laviosa 2002: 69).Dorothy Kenny (2001)  specifically refers to lexical normalisation throughout Lexis and Creativity in Translation.A Corpus Based Study; other related terms include simplification, explicitation and naturalisation.While there are some nuances in these words, they do not concern us especially in this paper, where the term normalisation is simply used to describe strategies that produce a fluent and conventional translation rather than a linguistically novel one.
Fluency in the target texts is rarely achieved through the complete elimination of a ST neologism; only a handful of omissions are present in total.Koelb removes schwergeschmückt, Lowe-Porter üppig-untauglich, Befallenen (with some compensation, another rarely used tactic) and Gepäckbeförderungsamt, the last of which is also omitted by Doege, as are breitschattend, bräunlich-ledern and Glücksfristtoo few to reveal any clear patterns, whether in terms of translators (i.e. which ones have a tendency to omit coinages) or translation strategies (i.e. when omissions take place).Generally, translators transform neologisms into conventionalised items, which come in any form imaginable and are greatly varied, for example, reducing creative words to their bare minimum with single unit items sometimes even at the cost of meaning (Dämmerblässe as paleness; Fremdenpoesie as song; Massenzutrauen as trust) to increasingly complex grammatical structures (such as established compounds, multi-part noun phrases and relative clauses).This observation applies to all categories of neologisms, although some distinctions can be made in terms of conventionalisation that takes place depending on the type of neologism concerned.

Hyphenated compounds
Spelling alternatives typically are translated with non-hyphenated compounds consisting of two nouns (e.g.Granatapfel-Getränk/pomegranate drink; Amethyst-Geschmeiden/amethyst jewelry) or noun phrases that are postmodified with prepositional phrases (Friedrich-Roman/novel on Friedrich; Maja-Welt/world of Maia).Although there is of course occasional variation with some items (Seemanns-Überjacke is rendered with a possessive in most TTs) and on the part of individual translators (Chase opts for "a chorus in which the entire ensemble laughed as hard as it could" for Lach-Refrain, in contrast to laugh refrain or laughing refrain chosen by others), the translations are generally straightforward, at least compared to NOM neologisms.These are sometimes rendered with simple constructions in the English versions (e.g. two adjectives or nouns joined by the conjunction and, a noun with a qualifying adjective), however, more extensive rephrasing is increasingly seen, as demonstrated by several of the translators -Burke, Lowe-Porter, Luke, Neugroschel, Heimin linking of two lexical units through hyphenation also means that the sense of the resulting item is not quite the same as that when the two units are interpreted separately, even more so as most combinations are unexpected and seemingly contradictory (as the just mentioned schwermütig-enthusiastisch illustrates).The separation of the lexical components thus implies a shift in meaning, which may be slight, but is nonetheless present as part of the normalisation process.

Non-hyphenated compounds
Non-hyphenated compounds show the same kind of fluency as other neologisms.While quite a range of different solutions are offered by translators, the use of noun phrases is prevalent with N + N compounds (no hyphen but a space between nouns, e.g.Farrengewucher / fern clusters; Gepäckbeförderungsamt / luggage office), ADJ + N phrases (e.g.Allerweltsferienplatz / cosmopolitan resort; Greisenlippe / senile lips) and N + of + N constructions (e.g.Glücksfrist / period of happiness; Raumeswüste / wilderness of space) dominating.

Creative variant of existing forms
The single ST example of a creative variant neologism is normalised in all TTs (Table 12).Two translators (Lowe-Porter, Doege) use a hypernym that results in a partial loss of meaning, while the remaining translators try to preserve the Halb-part of the original coinage by using a range of syntactic constructions, including AJD + NP, NP with a post-modifying PP and participle clauses, with all TTs confirming the preference for fluency.While neologisms that are derivations, conversions or creative combinations of multiple word formation processes are rendered creatively more frequently than compounds or the single creative variant example, normalisation is still the most dominant translation strategy.Fluent TT forms are diverse and, in contrast to other neologism types, demonstrate no clear preferences for a particular way of translating.Given the varied composition particularly of creative combinations, the absence of a dominant common pattern is, however, not so surprising: the divergent semantic and morphological complexity in items such as ausstürmend and (der) Weiterherkommende are more likely to require a wider range of translational solutions than some other neologism types.

Conclusion
While neologisms may not be a particularly dominant feature in Thomas Mann's Der Tod in Venedig, this makes them no less significant.Different types of neologisms are used in the novella, with compounds (both with and without hyphens) being employed most often, but conversions, derivations, a creative variant of an existing form and creative combinations involving multiple word formation processes all also featuring.As was to be expected, Mann's neologisms vary in their degree of innovation.Indeed, some neologisms, like SPA hyphenated compounds, may display stylistic preferences more so than linguistic creativity.The treatment of the ST neologisms in translation is, however, surprisingly uniform, given the large number of TTs and the extended time period of nearly ninety years that they span: they nearly always disappear as translators opt for normalisation.When neologisms are preserved in translationbetween 6.54 and 8.41 percent of the timeless innovation is present.TT coinages are, without exception, hyphenated compounds.These compounds, like Mann's, vary in their creativity.However, no translator experiments with any other word formation technique.Interestingly, the few coinages that are rendered creatively in most translations (i.e. by eight or more TTs), see the exact same form used by all but one or two translators, demonstrating less creativity overall than Mann.
The fact that all English Der Tod in Venedig translations approach neologisms in a similar manner raises questions about the so-called Retranslation Hypothesis that proposes, among other things, that early TT versions tend to domesticate while later ones increasingly foreignise.The data retrieved as part of this study so far show no such indication.It seems to corroborate those critics who have challenged the Retranslation Hypothesis and argue that retranslation is a complex phenomenon where "historical context, norms, ideology, the translator's agency and intertextuality" (Gürçağlar 233) must all be considered.As this paper focuses exclusively on ST neologisms in translation and also does not inquire into the background of how the different English Venices came to be and which factors may be influencing translational choices, additional exploration is needed in order to determine whether the uniformly domesticating approach applies not only for neologisms, but also other forms of linguistic creativity or even more generally for Mann's novella.Furthermore, it still needs to be investigated whether translators add any creative coinages of their ownindependent of the STthat may compensate for the normalisation of Mann's neologisms in other parts of the text.

Table 1 :
Hyphenated compounds in Der Tod in Venedig, according to type This kind of compounding, in which lexical items are linked by an en dash, is atypical in German but notably prevalent in Der Tod in Venedig with 43 examples (the full list is given in Table

Table 2 :
Non-hyphenated compounds in Der Tod in Venedig

Table 7 :
Creative translations by translator and neologism category

Table 8 :
Creative translations per neologism category and type

Table 9 :
Table 9) among them, as most translators use the exact same coinage.Neologisms with high TT preservation rates

Table 10 :
Einsam-Stumme neologism in TTs NOM neologisms are complex items.They provide a challenge due to nominalisation, which exists in English, but which, makes in comparison with German for a more marked (due to being a less productive word formation process) as well as more ambiguous word (due to absence of inflections).The third type of hyphenated compounds, DADJ neologisms, also manifests specific translation preferences.Hyphens are removed from the adjective unit, as it is virtually always split into two parts.Adjectives are often kept, although in some cases one component may be changed to an adverb or a noun.Punctuation or conjunctions may be added.Two representative examples can be seen with ängstlich-übermütig and bräunlich-ledern in Table11(below).DADJ and NOM neologisms are similar in some ways in that both involve the linking of two items with a hyphen that, except for word order, are equal: the components of the whole unit qualify each other, rather than one acting only as the modifier for the other.In translation this equality may be lost, particularly if one component is transformed into a different word category.The addition of conjunctions with some DADJ is also interesting, especially when but and yet are used as these verbalise contrast which is sometimes implicit (schwermütigenthusiastisch / melancholy yet enthusiastic, my emphasis) and sometimes not Chase's translation "the solitary and silent" illustrates: embedded in the clause "[t]he observations and chance encounters of the solitary and silent are more blurred" (my emphasis), the words can refer to either a person (a person that is solitary and silent) or a state of existence (solitariness, silence).The majority of translators clarify through grammatical changes, e.g.transforming the ST nouns into adjectives paired with an explanatory noun (one, man, person) or use other, more fluent solutions such as relative (B1 and B2, N) or participle clauses (LP's particularly extensive rendition).

Table 11 :
Translations for two typical DADJ neologisms