Germanic Vs Romance: An investigation into
the attitudes towards the languages of the
Germanic and Romance families and what
factors may influence them.
Christian Nicholas Mark Schramm
S3528588
MA thesis in European Linguistics
Departments of Applied Linguistics and Frisian Language and
Culture
Faculty of Arts
Rijksuniversiteit Groningen
Supervisors:
Dr. C.S. Gooskens
Dr. J.M. Fuller
4
thMarch 2019
Table of Contents
Abstract……….0
1. Introduction………...1
2. Background………...2
2.1. Language Attitudes………2
2.2. Pre Vs Post Attitudes……….7
2.3. Speaker and Participant Gender………...8
2.4. Participant Linguistic Background………8
2.5. Language Identification………...9
2.6. Romance Languages………..9
2.7. Germanic Languages………...16
3. Research Questions……….23
3.1 Hypotheses………25
4. Materials……….26
4.1. Procedures………29
4.2. Research Participants………...32
5. Results……….33
5.1. Post Ratings……….33
5.2. Pre Vs Post Ratings……….37
5.3. Pre Vs Post Ratings and Speaker Gender………39
5.4. Pre Vs Post Ratings and Participant Linguistic Background…………...42
5.5. Speaker Gender………45
5.6. Participants Gender………..47
5.7. Participant Linguistic Background………..48
5.8. Language Perception………....49
6. Discussion………...53
6.1. Overall Post Ratings………53
6.2. Germanic Vs Romance Languages………..61
6.3. Pre Vs Post Attitudes………...62
6.4. Speaker Gender………....63
6.5. Participant Gender………...63
6.6. Participant Linguistic Background………..64
6.7. Language Identification………...64
7. Conclusion………..65
7.1. Limitations………...66
7.2. Future Research………...67
0. Abstract
This paper reports the findings of an investigation designed to elicit the attitudes towards ten different languages originating from the Germanic and Romance families. 73 participants of all different backgrounds took part in an online survey, where they had to listen to recordings of these languages and rate them on several different scales. As well as exploring the attitude scores given by the participants, some factors that may influence them were also looked at. These factors included attitude scores given pre and post listening to the recordings, speaker and participant gender, participants linguistic background and how easy a language was to identify. Results showed that English was the highest rated language, with Romanian finishing as the lowest. Additionally, pre and post scores differed significantly for English, Dutch and German. Male participants were also found to prefer female voices to their own, while females rated their own voices higher. Some correlations could also be made between a language being harder to identify and it receiving lower attitude ratings. The findings were used to try and give more current attitude ratings towards these languages that can be used in further research. Reference was made throughout to language attitudes and the theories and hypotheses that underpin it, as well as the historical development and structure of each language that featured in the investigation.
1 1. Introduction
In Europe there is a vast amount of language variation, with many different languages being employed across the continent. These languages each have long and varying histories, which has not only affected their structure and characteristics, but also the attitudes that people hold towards them. In this investigation, ten languages from the Germanic and Romance language families will be assessed, with the aim of finding out modern day attitudes towards these languages and what may influence them. The languages will be assessed by participants in an online survey that has been designed to explore language attitudes from a variety of different perspectives. Five languages from each family were selected based on them being official languages of the European Union, they are as follows:
➢ Romance: French, Italian, Portuguese, Romanian, Spanish. ➢ Germanic: Danish, Dutch, English, German, Swedish.
These language families have been chosen as although they both originate from Indo-European, there has been much change and divergence over time, which has even led to languages from within these families becoming mutually unintelligible. Additionally, the Romance languages all derive from Latin, a highly prestigious and historical language, spoken mainly by the educated, higher class (Penny 2002). However, Germanic languages derive from Proto-Germanic, which started in the Baltic region of Europe and carried little or no prestige as a language (Henriksen and Van der Auwera 1994). Taking an example on how attitudes towards the languages in these families may differ, research by Giles and Niedzielski (1998) states that history indicates that German is perceived as ugly and grotesque, while French is found to be a beautiful and flowing language. Because of this imbalance in power between the two language families, Romance languages like Latin and French were able to have a large impact on several Germanic speaking nations, such as England and The Netherlands and it was only resistance from staunch locals that kept their original Germanic tongues intact. Since then, language development, interaction with other languages, and our expanding modern-day society has allowed for historical ideas of languages to develop overtime, meaning historical opinions of Romance and Germanic languages may have ameliorated or prejorated over time.
2
So far, reference has been made to attitudes of well-known languages, such as English, French and German. However, attitudes to languages such as Danish, Swedish, Portuguese and Romanian are rather less studied. Considering this, it will be very interesting to see how attitude ratings towards these languages will score. Additionally, as well as exploring attitudes towards each language, factors that might influence these attitudes will also be assessed. This includes participants ratings pre and post listening to recordings of each language, the influence of speaker and participant gender, participant’s linguistic background, and how easy a language is to identify amongst others.
To begin, some theoretical background on language attitudes will be provided, as well as more information on the aforementioned factors that may influence language attitudes. Each language will then be discussed, including current or historical attitudes they may hold. Following this, the research questions and hypotheses will be outlined, before an explanation of how the study was conducted. Results will then be presented with graphs and statistics, before then being discussed further in the discussion. Finally, conclusions, limitations of the study and ideas for future research will be outlined.
2. Background
Some background on language attitudes, how to elicit them and what might influence them will now be provided. Moreover, some factors that may impact the language attitudes given in this investigation will also be introduced.
2.1 Language Attitudes
Crystal (2008: p. 266) describes language attitudes as “the feelings people have about their own language or the language(s) of others”. When trying to understand why these attitudes may be formed, Garrett (2010: p.22) stated that we learn attitudes in a variety of different ways, and that “Two important sources of attitudes are our personal experiences and our social environment, including the media”. Expanding on this, people hold two kinds of attitudes; implicit and explicit. Explicit attitudes are rather clear, with Breckler et al. (2006) reporting that they are attitudes than can be given consciously by an individual. However, these attitudes sometimes tend to not be completely representative of an individual’s opinion, which is where implicit attitudes come in. Implicit attitudes are defined by Breckler et al. (2006: p.202) as
3
“Automatic evaluative responses to a target, which may occur without awareness”. This is expanded on by Greenwald and Banaji (1995) who suggest that implicit attitudes seem to be defined more by past experiences which may incite positive or negative feelings towards something. In a research context, it can be difficult to differentiate between implicit and explicit attitudes, with Wittenbrink and Schwarz (2007) explaining that in attitude tests, participants often have time to reflect on their responses and will give more explicit attitudes biased towards a certain answer, or unrepresentative of their genuine attitudes. Beckler et al. (2006) add to this by explaining that implicit and explicit attitudes can sometimes be inconsistent with each other. Using an example that might be specific to this investigation, someone may have a negative implicit attitude towards French, because of a previous experience with the language. However, they may explicitly rate it highly, as general opinions towards the language are high and they may not wish to offer a differing opinion to this. Some methods to elicit more implicit language attitudes have been designed, for example, the matched guise technique (Lambert et al. 1960). Matched guise uses speech samples of a proficient bilingual speaker amongst recordings of other speakers, so participants are unaware that they are listening to the same speaker using two different languages. Following listening to the speech samples, participants are asked to give their attitude ratings of the language. Should there be any differences in attitude towards the two languages of the same speaker, it can reveal indirect language attitudes, since language is the only thing that differs between the two recordings (Lambert et al. 1960). For this investigation, matched guise could not be used, since it was impossible to find people who spoke all ten languages to a proficient level. More information on how this study aimed to differentiate between implicit and explicit attitudes will be given later.
Having looked at types of attitudes, reference will now be made to what may impact and influence them. This can be done by looking at Giles et al’s. (1974) research, which specified two different hypotheses that influence language attitudes. The first of these is the inherent values hypothesis, which suggests that certain varieties and languages can be regarded as more prestigious than others due to having more pleasant inherent characteristics. While there’s little research to support the claim that some languages are inherently more beautiful than others, Crystal (1995), did suggest that there could be certain characteristics of sounds that make them more pleasant to listen to. Using English as an example, Crystal (1995) suggested that consonants pronounced in the front of the mouth are deemed as more pleasant than ones pronounced from the back. The same can be suggested for vowels, implying that fewer back sounds leads to a language being perceived as more beautiful. Looking at an English example,
4
the word ‘pretty’ has a short, clipped front vowel sound, while ‘ugly’ has a broader, heavier back vowel sound. Looking at the meanings of these two words, this perhaps illustrates the difference in aesthetics between frontal and back vowels.
Looking at some other languages sounds, Danish could be seen as less beautiful due to the presence of theStød, which represents an uncomplete glottal stop, a sound that is difficult to reproduce and deemed rather ugly (Haberland 1994). While each language has their own unique vowel and consonant sounds, it is also interesting to see how these differ with other languages. For instance, the consonant cluster ‘sch’ in Dutch, which is akin to a fierce ‘g’ sound pronounced in the back of the mouth, was used by Dutch authorities to identify German spies, as the ‘sch’ in German has a much softer and more frontal pronunciation, like ‘sh’ in English (Van Koppen 2014). From these examples, we can see how sounds in the back of the mouth could be deemed as ugly, not just in isolation, but also in comparison to how other languages produce these sounds. Additionally, these examples are quite rare sounds that are exclusive to these languages, which immediately puts them at risk of being negatively seen, as they are not familiar to speakers of other languages (Deutscher 2010). Deutscher (2010) adds to this by saying that Italian, which is generally considered a beautiful language has very few sounds that do not feature in other European languages. This raises the idea that the more familiar a language is, the more positively it may be rated. At this point, it must also be considered just how subjective language attitudes can be, with Giles and Niedzielski (1998) explaining that while certain factors like phonological features of a language can influence someone’s opinion, final assessments are placed firmly in the ear of the beholder. In this study, the final ratings given to languages, coupled with their phonological features and familiarity can be compared to see if they may impact participants ratings.
The second hypothesis outlined is that of imposed norms, which is strongly based on stereotypes, cultural norms and the consensus of other people. Giles et al. (1974) reported that many language attitudes today seem to have been formed by historical cultural norms, which have been firmly established and contribute to present-day stereotypical views of a language or variety. Using the case of English, Giles et al. (1974) comment that had high-end institutions, such as courts been set up in different areas of the country, the accent or variety of that area would have then gone on to receive the highest prestige and gravitas. This is interesting, as it shows how language attitudes can be formed solely from cultural norms and familiarity, with the ability to continue for generations.
5
Looking at other examples and expanding on previous findings of Giles and Niedzielski (1998), French is often cited as a beautiful and prestigious language, with it rivalling and eventually overtaking Latin as a language of law and administration, as well as influence, prestige and power (Battye et al. 2000). Additionally, since 1635 the Académie Française has been established to help preserve, standardise and guard French from linguistic shift and change. While French’s history would seem to indicate attitudes towards it will be positive, the same cannot be said for German, with Liebscher and Dailey-O’Cain (2013), explaining that negative attitudes towards the language today may still index the Nazi period, potentially more so amongst the older generation. Studies from Dekker et al. (1998) support this, finding that Dutch students learning German in school still allowed the Second World War to cloud their judgments of the language, people and country, seeing it as overpowering, arrogant and superficial. Finally, as previously mentioned, the Romance languages all derive from Latin, which was a highly prestigious language. During the development of the Romanian language, in response to the growing influence of the Slavic languages, a period of Re-Latinisation took place to bring the languages prestige closer to that of the other Romance languages and eliminate ‘ugly’ or ‘simple terms that had entered the language from the Slavic languages (Hanners 1999). This desire to be closer to Latin highlights how an imposed norm of Latin being a beautiful language to aspire to has altered the development of certain languages, and perhaps also the attitudes towards them. The following examples highlight just how much history and stereotypes of a language may continue to persist, resisting change and development. In the context of this study, it will be interesting to see if ratings of the language point more towards their historical past and opinions or be more reflective of a more open modern-day society.
Trudgill and Giles (1978) also devised a third hypothesis of social connotations, which focusses more on the cultural and social prestige of a language linked to where and how it is spoken. Trudgill and Giles (1978) initially use the example of English, stating that Received Pronunciation (RP) was seen amongst UK natives as the nation’s most beautiful variety, being linked with class and prestige, followed by regional accents and then City accents. Decades on, this opinion seems to persist, with Rindal (2015) suggesting that RP is viewed as the most proper way of speaking English, due to its close connotations with wealth and status. Trudgill and Giles (1978) also add that regional accents are more positive than City accents in the UK as they are associated with a more pleasant lifestyle and environmental placing. Studies by Van
6
Bezooijen (1994) into the Dutch language also associated regional varieties with friendliness and linked to professions such as a nurse, or farmer’s wife. Looking at British City accents, such as ‘Brummie’, from Birmingham, various studies have indicated that UK natives find the accent to be untrustworthy, ugly and stupid (Dixon et al. 2002; Crystal and Crystal 2014). Comparing RP and Brummie, RP is traditionally spoken in the South-East of the United Kingdom, where London and more affluent areas of the country are located, while the City of Birmingham is traditionally a very industrial and working-class area of the country. This presents the argument that the socio-cultural region where an accent is spoken heavily impacts the attitudes towards it. Van Bezooijen (1994) also presented the idea that greater intelligibility of a language often leads to more positive attitudes, suggesting that languages or accents with higher comprehensibility are more favoured. This will be interesting to consider when we look at the languages in this investigation, as participants from certain areas or social groups may see certain languages or the accents of the speaker as more beautiful, due to increased familiarity and intelligibility.
Using an example from one of the languages that feature in this investigation, the origins of the Standard Italian language we recognise today is from Florentine, a Tuscan dialect spoken in Florence. Like RP, Florentine was viewed favourably by the upper-class and was also the language of choice of esteemed writers, such as Dante, Petrarch and Boccaccio (Maiden 2013). That Florentine was such a prestigious variant and grew to form the initial structure of Standard Italian highlights how social and cultural prestige can impact language growth and development, as well as attitudes. This investigation will aim to explore whether the connotations that have previously existed towards other languages are still considered true, or whether languages of different origins and styles are also favourably rated in comparison to historically very prestigious or beautiful ones.
Having now seen how language attitudes can vary in a significant way, this highlighted the investigations need to include a broad range of questions that would assess different attitudes. In order to do this, Zahn and Hopper’s (1985) category system for assessing language attitudes was used. These categories are dynamism, attractiveness and superiority, all chosen as they are often considered when assessing a languages beauty. Under these categories, come several different opposing adjectives, for example ‘beautiful’ and ‘ugly’, which would come under the attractiveness category. This system is an effective way of eliciting attitudes and allows speakers and the language they use to be assessed from a variety of different stand-points. For
7
example, while a speaker may not be viewed as ugly, they could be smart, kind and normal. Contrarily, the speaker may be perceived as beautiful, but stupid, unkind and strange. For this reason, speaker and language attitudes cannot be viewed as straightforward, but must consider many different factors for the speaker and their language to be assessed fairly and as objectively as possible. Garrett (2010: p.2) highlights this by explaining that people hold language attitudes on a variety of different means, such as “Spelling and punctuation, words, grammar, accent and pronunciation”. These linguistic factors will differ between each language, making the attitudes towards them interesting to consider in isolation, as well as in comparison with each other. Having discussed some hypotheses towards language attitudes, some additional factors that could determine people’s overall attitudes towards the languages in this investigation will now be considered.
2.2 Pre Vs Post Attitudes
Looking at the hypotheses of Giles et al. (1974) and Trudgill and Giles (1978), it is clear to see that there are many pre-existing attitudes and ideas surrounding languages, some of which have been given in section 2.1. For this reason, the investigation will be set out in such a way that participants will be asked to give their ratings of a language before listening to the recordings of it, and twice more after listening to the two recordings of each language. This is done with the belief that participants ratings, pre-listening to the recordings will be based more upon the imposed norms and social connotations linked with the language, while the ratings given post listening to the language will reflect more genuine opinions based on how they thought the speaker sounded, which will be more implicit and reflect more inherent values towards the language. The ratings given post listening to the recordings will be especially interesting as if they differ from the pre-given opinions, then we can suggest that the opinions of the participants were more affected by stereotypes and social norms, rather than being concerned with the features of the language. For example, while a certain participant may go into the investigation holding a pre-conceived attitude of German as an ugly, harsh and arrogant language, after listening to recordings of it, they may realise that their opinion is not as negative as they imagined and give it a more favourable rating. This can be assessed in the other direction too; perhaps French is not as beautiful or prestigious as people first thought, with post attitudes scoring lower than pre-given ones. The pre and post attitude comparison will not only show us which languages may be affected most by negative stereotypes, but also perhaps give more current and accurate opinions towards each language, which are less impacted by Giles et al
8
(1974) and Trudgill and Giles’ (1978) hypotheses, but more on the many linguistic factors that were outlined by Garrett (2010: p.2).
2.3 Speaker and Participant Gender
Looking at speaker and participant gender, research into potential speech differences has been made previously, with Lakoff (1975) suggesting that males are more careless with their speech, while females use more standard terms and fewer rude or aggressive terms. However, as all speakers in the investigation will be reading the same text, analysis into how the two genders communicate will not be made, rather, the properties and style of their voice. According to Trouvain and Zimmerer (2017: p.2238), voice plays a key role in the evaluation of “beauty and pleasantness, sympathy, politeness, competence, persuasion and charisma”, certain factors that affect this include, “pitch and pitch range”, “vowel quality” and “voice quality”. However, it is largely unknown whether this differs greatly amongst genders, therefore, ratings for male and female voices will be considered and compared to see if any factors influence participants ratings. This could be according to the factors outlined by Trouvain and Zimmerer (2017), however, with just one male and female speaker for each language, this cannot be compared across languages. Although, with ten male and female speakers respectively, we can check for differences in ratings between the two genders and search for initial ideas on which genders voice may be preferred.
As well as speaker gender, participant gender will also be investigated to see if there is any difference in how males and females rate each other’s voices. This will refer to whether there is any preference amongst males and females for a certain genders voice. According to Trouvain and Zimmerer (2017) adult hetero-sexual listeners find voices of the opposite sex more appealing than that of their own. Without knowing participants sexual orientation, this idea can still be tested to see if there is a preference amongst males and females for voices of the opposite gender, while a preference for the voice of one gender being preferred should not be discounted.
2.4 Participant Linguistic Background
Another factor that should be carefully considered is the linguistic background of the participants. As the participants came from all over the world, some different groups depending
9
on this were made. Looking again at Giles et al. (1974) hypothesis of imposed norms, as well as Trudgill and Giles’ (1978) hypothesis of social connotations; cultural and stereotypical viewpoints do influence language attitudes and someone’s linguistic background may affect that even further. For example, it could be that participants of a certain nationality or linguistic background may see other languages more negatively than others due to socio-cultural, historical and geographical reasons. Some specific examples are highlighted in section 2.1, such as Dekker et al’s (1998) research of Dutch students then negative opinions towards the German language that referenced the Second World War. While this is an example based more upon nationality, rather than linguistic background, reference can be made to the fact that Romance languages are looked upon favourably, due to their links with the Latin language, which could suggest positive attitudes towards the Romance languages from participants of all linguistic backgrounds. For the Germanic languages, while attitudes to the German language seem negative, more favourable attitudes of English, as well as the opinions of Swedish, Danish and Dutch, suggest that ratings of the whole language family may be more varied. It will also be interesting to note how speakers of Germanic languages rate Romance languages and vice versa, seeing if they prefer languages that are similar in nature to their own, like Van Bezooijen (1994) suggests, or ones from a different family that have varied linguistic features.
2.5 Language Identification
In the investigation, it is expected that participants will have some difficulty correctly identifying certain languages. For example, Swedish and Danish share many similarities and can easily be confused, while other languages, such as Romanian and Portuguese are not as mainstream as English or Spanish. However, should languages have a large amount of incorrect identifications, it could mean that there is a correlation between identification of language and its final ratings. For example, it could be that languages that are difficult to identify hold more negative opinions as they are more unknown to participants. While more familiar languages will receive high opinions as they are more well known amongst the participants. This agrees with the findings of Deutscher (2010).
10
Reference will now be made to the history, development, characteristics and current attitudes towards the ten languages that will feature in the investigation, starting with the Romance languages.
Italian
Looking firstly at Italian, Grandgent (1927: p.9) described it as “the most conservative of the neo-Latin tongues”, closest in style to Classical Latin. Italy is a very interesting linguistic region to consider, due to the presence of its many dialects, which were born from the spread of Latin and the Roman Empire (Maiden 2013). Lepschy and Lepschy (1988) give more explanation to the variety found amongst these dialects by specifying that an Italian ‘dialect’ refers to a spoken and written variety, different to Standard Italian. In fact, Maiden (2013) explains that the Italian dialects we recognise today cannot be considered descendants of an Italian language, rather, the Standard Italian language grew from one of these specific dialects. As mentioned in section 2.1, Florentine was the dialect that emerged as the most prestigious variety and would eventually form the origin of the Standard Italian we recognise today. However, it would still take centuries for a standard form of Italian to break through and unite Italy linguistically. In fact, post-unification of Italy in 1861, only 3% of the population spoke what was perceived Standard Italian at the time, with the population’s majority communicating in their respective dialects (Alonzo 2012). Even following World War Two, standard Italian was still only spoken by a third of the population (Alonzo 2012). However, moving towards the 21st Century, growth in media, exposure and the need for a national standard led to Italy unifying linguistically. Despite this, Italy is still regarded a country of dialects, with many of these differing from each other far more radically than in other European Countries. For example, Aikhenvald (2015) explained that a speaker from Rome would not immediately understand someone from Sardinia, without having to learn some aspects of their dialect. A further example of dialectal difference in Italy, including Sardinian, is illustrated in table 1.
Table 1. Lexical differences between Italian dialects.
English Italian Sardinian Salentine Emilian Friulian
Hen Gallina Pudda Caddina Galeina Gialine
From an attitude perspective, Latin’s prestigious history, as well as Florentines usage by the educated elite would suggest Standard Italian would be a beautiful language in the present day,
11
especially when considering its ongoing close links to its original varieties. A study from Gooskens and Van Heuven (2019), investigated language attitudes towards Italian of other Romance language speakers and it scored consistently above 4, in a Likert-scale, which scaled from 1, ‘ugly’, to 5, ‘beautiful’. Considering the high attitudes of Romance language speakers towards Italian, it will be interesting to see how speakers of other languages perceive the language. Additionally, due to the strong prevalence of dialects and varieties in Italy, it could be that the speakers chosen for this investigation will have different accents and manners of speech that impact results.
French
French’s linguistic origins began with the Roman invasion of Gaul, which was situated in the whole area of what is known as France today, as well as small parts of Belgium, Switzerland and Italy (Rickard 1989). At the time a Gaulish language was spoken, however, Latin soon became imposed on the population, which saw a change in alphabet and the prestige of the Gaulish language drop remarkably. In fact, Adams (2003) reports on the fact that Gauls used to Romanise their names and use Latin endings to create more prestigious identities of themselves. Eventually, by the end of the 5th Century, the majority of Gaul were used to speaking a Latin tongue, although there was some Gaulish influence, with many loan words as well as pronunciation impacting the language in such a way that it could be differentiated from Classical Latin (Rickard 1989). Following the diminishment of the Roman Empire, Germanic invaders were next to arrive, however, rather than imposing their Germanic tongues on the people, they grew to adopt the Gallo-Roman dialects that were spoken in the areas they settled in (Nadeau and Barlow 2008). Battye et al. (2000) suggest that the reason behind this is that they saw the Gauls and their language as more sophisticated than theirs. Despite this, there was still a large amount of Germanic impact on these Gallo-Roman dialects, lexically, syntactically and phonologically. Following the settling of the Germanic invaders, two main languages came to power; ‘langue d’oil’ and ‘langue d’oc’. This was an interesting period for France, as they began to export their languages, before even having a national standard themselves. Langue d’oc made inroads across the Mediterranean, while langue d’oil famously made its way to England, where it quickly imposed itself as a language of high prestige (Nadeau and Barlow 2008). Back in France, due to langue d’oil being spoken in Paris and the rise in power of the Franks, it started to overcome langue d’oc and emerge as the nation’s standard (Nadeau and
12
Barlow 2008). Some examples of Modern-day French (langue d’oil) and Occitan (langue d’oc) are illustrated below.
Table 2. Lexical differences between French and Occitan.
English Welcome Street Afternoon Box
French Bienvenue Rue Après-midi Boîte
Occitan Benvengut Carrèira Vespre Caissa
Following langue d’oil’s emergence as the national standard, French continued to develop swiftly and efficiently (Rickard 2003), with Nadeau and Barlow (2008) claiming that in an English dominated world, French has grown to be a de facto second language, being the ninth most spoken language in the world, but also having official status in over twenty countries and being widely studied as a second language. In fact, amongst many native Anglophones, French is seen as a language of “chic, taste and superiority” (Nadeau and Barlow 2008). Additionally, looking again at Gooskens and Van Heuven’s (2019) research, French was rated as very beautiful by other speakers of Romance languages, reaching scores an average of 4 out of 5. This, coupled with the favourable views of Anglophones, would suggest that French will receive high ratings befitting of its historical stature, prestige and class.
Spanish
The Latin origins of Spanish first came about when the Roman Empire spread to Spanish territories and imposed their language on the people (Penny 2002). However, like with French, the diminishing of the Roman Empire led to new varieties of Latin being spoken in the region (Pharies 2015). Penny (2002) elaborates on this by explaining that this was mostly thanks to the Visigoths, who spoke Latin bilingually with their own East Germanic tongues. Although, this did not signal a truly significant amount of linguistic change, with them eventually dropping their bilingualism in favour of Latin, due to them seeing the early Spaniards culture as superior to theirs (Penny 2002). It was only until the arrival of Islamic invaders in the 7th Century that notable linguistic change took place, with a huge amount of lexical and semantic borrowing taking place from Arabic (Penny 2002). Having control of three quarters of the peninsula, it was only until the reconquest of Spain by native Spaniards that ousted these controls and led to the establishment of the Kingdom of Castile in the 10th Century (Pharies 2015). The Castilans spoke what Penny (2002) describes as a rather bizarre variety of Hispanic
13
Latin, which gained significant cultural and territorial spread during the reconquering period. This culminated in the liberation of Toledo in the centre of the country, which signalled great linguistic change, and by the 12th Century, Castilan had continued its growth, with its first written documents being produced (Pharies 2015). Following the uniting of the kingdoms of Castile and Aragon, Castilan was able to easily reach out to the rest of the peninsula and although minority groups continued to speak their respective languages, such as Basque or Catalan, Castilan spread as the national language. Basque and Catalan are still spoken today in Spain by rather large minorities of the population, differing greatly to Castilan, or what we know as Standard Spanish, some examples of this are illustrated below.
Table 3. Lexical differences between Spanish, Basque and Catalan.
English Spanish Basque Catalan
Bird Parájo Txori Ocell
Want Querer Nahi Voler
Eat Comer Jan Menjar
Morning Mañana Goizean Matí
As well as the Kingdom of Castile coming to power, the Castilan languages growth also has much to do with becoming the main language of literature, imposing itself prestigiously and powerfully, with King Alfonso X appointing a team of experts to produce books on matters concerning law, church, marriage and inheritance, all in Castilan (Nadeau and Barlow 2013). In fact, he was one of the first European Kings to highlight the need for standardisation of their vernacular and during his reign, texts written in Castilan went from having a great deal of variation to almost none (Nadeau and Barlow 2013). As well as spreading in Spain and continuing its growth, Spanish also continued its spread globally, travelling to America, the Mediterranean and the Philippines (Penny 2002). Today, Spanish is spoken by the vast majority of South and Central America and can be put in the same camp as French and Italian when considering ‘beauty’ of languages. According to Gooskens and Van Heuven’s (2019) research it is very well-rated amongst other speakers of Romance languages, receiving scores of almost 4 out of 5.
Portuguese
With Portugal and Spain sharing the same peninsula, their linguistic histories share a great deal of overlap. During the previously mentioned 7th Century Islamic invasion, many differing
14
vernaculars were spoken, with the West of the peninsula, where Portugal lies today, speaking a language described as Galician-Portuguese. This combination would form the basis for the Portuguese language that we know today (Azvedo 2005). Although Portuguese split with Galician, the language is still spoken today, but only by a minority in the very North-West of Spanish territory. In fact, the languages share very high mutual intelligibility and it was only until Portuguese borders became more defined that a Portuguese culture, identity and language began to diverge from Galicia (Azvedo 2005). This was initially a steady process, which gained pace in the Middle Ages, when Romance languages started to gain their own identity and commence their processes of standardisation. This resulted in Portuguese reshaping itself, ridding itself of constructs deemed too close to Galician and directing itself more towards the Portuguese spoken in the Southern area of the territory (Azvedo 2005). Portuguese at the time gained more of an identity and ended up being unchallenged in becoming the official language of the Portuguese territory of the peninsula. Some differences between modern-day Portuguese and Galician are shown in table 4.
Table 4. Lexical differences between Portuguese and Galician.
English Thank you Monday Young Forest
Portuguese Obrigado Segunda-feira Jovem Floresta
Galician Grazas Luns Novo Bosque
Once Portuguese began to establish itself as the dominant language in Portugal, it also spread to a variety of different African nations such as Cape Verde, Angola and Mozambique, and perhaps most notably, to Brazil in South America. Indeed, it is thanks to Portuguese’s status as Brazil’s official language that it is one of the most spoken languages in the world (Azvedo 2005). Looking at attitudes, due to its geographical location and similar history, Heeringa et al. (2013) explain that Spanish and Portuguese share many similarities both lexically and orthographically. At first glance, this would suggest Portuguese would have similar ratings to Spanish, however, what has not been considered yet is the languages distinct phonological differences. Jensen (1989), cites Ellison and Andrews (1969) assertion that although Portuguese and Spanish do share many similarities, phonology is not one of these and they have a surprising amount of difference in sounds. Knowing this, it will be interesting to see how much attitude ratings towards the two languages differ, especially having considered Crystal’s (1995) research on how certain sounds can be interpreted. Additionally, although it may not have as many speakers as Spanish, it does still have a large global influence, being
15
spoken in 4 different continents. Looking at Gooskens and Van Heuven’s (2019) recent research, Portuguese did not have as high opinions amongst other Romance language speakers, with an average closer to 3 than 4, so it will be interesting to see if this study elicits similar findings.
Romanian
Romanian is one of the most distinct Romance languages, due to its geographical location and positioning in the Eastern-Romance family, which derives from Vulgar Latin (Heeringa et al. 2013). Consequently, Romanian shares fewer lexical, orthographic and phonetic similarities than with the other Romance languages. Despite this, its base is still Latin, due it being occupied previously by the Roman Empire (Thompson 2015). Following the empires diminishment, Romanian began to diversify upon encountering their Slavic neighbours, while it also gained further influence from languages, such as, Hungarian, Turkish and Greek (Thompson 2015). Slavic languages were to have the most significant impact however, with Slavs being the first people to engage fully with all Romanians North and South of the Danube river. Although, unlike the Roman Empire, the Slavs did not force their own languages upon the Romanians and adopted bilingualism in both Slavic and Latin. However, the Slavs did choose to change many features of the Romanians language, with them substituting many Latin words, for their preferred Slavic terms, particularly for emotional terms (Hanners 1999). Table 5 highlights an example where the word ‘Love’, from Latin ‘Amore’ shows little variation amongst the Romance languages, while Romanian adopts a completely different form, which was taken from Slavic (Hanners 1999).
Table 5. 'Love' in all Romance languages.
French Italian Spanish Portuguese Romanian
Amour Amore Amor Amor Dragoste
Vocabulary was not the only thing the Slavs had an influence on, with pronunciation also differing greatly. This is because Latin was their second language, which they spoke with heavy Slavic accents. Resultantly, some Slavic elements of pronunciation were brought into the language, impacting tone, intonation, pronunciation and stress (Hanners 1999). However, as Slavic influence continued to grow, there were certain voices of dissent that felt the Romanian language was being taken away from them and that Slavic languages were beginning to have
16
too much influence politically, religiously and culturally (Boia 2001). In fact, a book written on Romanian etymology at the time had revealed that the language was two-fifths Slavic, one-fifth Turkic and just one-one-fifth Romanian (Boia 2001). This is when the period of Re-Latinisation, referenced by Hanners (1999) in section 2.1 took place. Romanian today can be understood as a huge mix of languages and influences, maintaining many of the Slavic terms and pronunciation, as well as continuing to be influenced by several languages that are spoken by its borders, such as Hungarian, Bulgarian and Ukrainian. Interestingly, in the Republic of Moldova, where Romanian is spoken, the language can even be written in the Cyrillic Alphabet (Thompson 2015). As well as this, Romance languages, particularly French, also influence the language, while English terms have also become popular (Hanners 1999). Romanian’s ratings in Gooskens and Van Heuven’s (2019) research were less favourable compared to the other Romance languages, with it falling below an average of three. Considering this, it will be interesting to see what speakers of other languages may think of Romanian, especially knowing how Romanian’s linguistic past has influenced it in a much different way to the other Romance languages.
2.7 Germanic Languages
English
English has undergone significant change over time, developing from a spoken variety with little prestige, to today’s Lingua Franca (Crystal 2005). Since its beginnings, English has been influenced and shaped by a vast amount of different languages, including Celtic, Latin and Germanic varieties of Indo-European languages (van Gelderen 2014). English’s first major developments came with invaders from Norway, Sweden, and The Netherlands, who arrived at different parts of Britain’s East coast, bringing with them a variety of different Germanic tongues (Hogg 2006). This began the period of Old English with the Germanic people, as well as their languages quickly spreading their influence further inland (van Gelderen 2014). This led to Germanification of the English language, including a strict and complex inflectional system of grammar, phonetic spelling style and many new vocabulary words (Burnley 2000). While English continued to gain influence from a vast array of different languages, it’s next major influence would be from French, which grew to power in England alongside William of Normandy’s rise to the English throne (Hogg 2006). In fact, the arrival of the Normans did not signal just linguistic change, but changes in aristocracy, authority and power (van Gelderen
17
2014). At this point, a period known as Middle English began, during which English kept its Germanic base, but saw the arrival of an abundance of French loanwords (Shay 2008). Previously, English had been described as a rather “vulgar” and “barbarous” language (Jucker 2000: p.43), however, more French terms allowed the language to evolve and gain more prestige. Table 6 illustrates the impact of French, with some examples of doublets. These are words that have one meaning in one language, which are then adopted by another language with an ameliorated meaning or idea (van Gelderen 2014).
Table 6. Doublets taken from French in the English language.
French origin English adaptation Meaning
Maison (House) Mansion Large, expensive house
Chef (Cook) Chef Professional cook
Palais (Dwelling) Palace Large residence for royalty
English’s next developments would come from within the country, with the invention of the printing press by William Caxton and the great vowel shift encouraging the beginning of a standardisation process that would take English into its early modern stage (Hogg 2006). These centuries also saw British colonial power spread the language across the Globe, whereby English become an official language in a vast number of countries spanning all 7 continents. Crystal (2008) comments on this, stating that by the end of the 20th Century, English had imposed itself so much that it became seriously considered as the World’s first Global Lingua Franca. The map below outlines the amount of states where English holds official language status:
18
Figure 1. States where the English language has official status.
Looking at attitudes to British English, which was the variety used in this investigation, research by Jenkins (2009: p.27) described the language as “clear”, “elegant” and “easy to understand”. While previous studies by Trudgill and Giles (1978) and Rindal (2015) have found RP to be a highly favoured variety of English amongst native English people. Alongside this, Gooskens and Van Heuven’s (2019) research found English to have a beauty rating of around 4 out of 5, amongst speakers of Germanic languages. Knowing this, it would seem as if English will gain positive attitudes from participants, in line with recent research and historical viewpoints. However, it is intriguing to consider how much of this is may be thanks to the large influence of the French language that changed many aspects of present-day English, as well as the fact that it is the Lingua Franca.
German
Because German territory was never breached by the Roman Empire (Sanders 2010), much of the German language we know today has received less influence from other languages. In fact, this explains why Germany, the Netherlands, much of Belgium and Scandinavia all kept their Germanic tongues, as opposed to having Latin imposed on them (Sanders 2010). Early on, these Germanic dialects varied depending on the location they were spoken, and it mainly remained a spoken variety, with little written evidence of early Germanic (Young and Gloning 2004). Additionally, early Germanic was not just limited to the Germanic nations, landing in France, England and Italy, which is also when German began to differentiate itself from other
19
Germanic dialects, such as Dutch, Frisian and Danish (Sanders 2010). A defining moment for the German language was a sound shift that separated the language into ‘High-German’ and ‘Low-German’, which brought about some major differences in pronunciation between North and South Germany (Sanders 2010). These differences would continue to develop and going into the early High-German period, the two varieties could be easily differentiated. During the early High-German period, another defining moment occurred, with Martin Luther’s translation of the Bible into German. According to Salmons (2018), when looking at the original German text and modernising it, there is little or no translation required, which encourages the widespread idea that German has not developed much over the years thanks to a long and stringent process of standardisation. However, German is a language with a significant amount of dialects and varieties in Germany, while its also spoken as an official language in Austria, Switzerland and Luxembourg. In fact, Luther’s translation may have been written in a way that all German speakers would understand it, however, it doesn’t resemble a specific form of standardised German, with many different styles and varieties existing across Europe (Clyne 1995). Indeed, it is only until recently that any form of general standardisation has taken place, and Germany has retained much of its dialectal diversity due to its late unification, while Austrian German, Swiss German and Luxembourgish have retained their own individuality, with the latter two even being recognised as their own languages (Clyne 1995). An example of the variation between German varieties and languages is shown in table 7.
Table 7. Lexical differences between German, Austrian German, Swiss German and Luxembourgish.
English German Austrian
German
Swiss German Luxembourgish
Good bye Auf
Wiedersehen
Auf
Wiederschaun
Widerluege Äddi
Wardrobe Kleiderschrank Kasten Chaschte Kleederschaf
Bread roll Brötchen Semmel Weggli Brautroll
Potato Kartoffel Erdapfel Härdöpfel Gromper
Looking at attitudes towards German, Giles and Niedzielski (1998) report that German is a language that is generally perceived as harsh, severe and unfriendly sounding. Moreover, in a study by Batram (2010: p.97), which investigated Dutch and English attitudes towards learning German, some students stated they didn’t like learning German due to its aesthetics, with adjectives such as “harsh”, “horrible” and “stupid” being used. Negative attitudes that reference
20
the Nazi period and Second World War have also been mentioned by Liebscher and Daley-O’cain (2013) and Dekker et al (1998). Moreover, looking at Gooskens and Van Heuven’s (2019) study, German had an overall attitude rating of under 3 amongst speakers of Germanic languages, which suggests rather ambivalent attitudes towards it. This study will take into consideration the opinions of participants of a different linguistic background, which could negatively influence this rating further.
Dutch
Compared to English and German, Dutch’s growth as a language was rather slow, with the Old Dutch period being difficult to analyse due to very few texts being written during this time, and variation of the spoken variant of Old Dutch being vast (Willemyns 2013). In fact, it was only around the 12th Century that Dutch became more prominent, with texts written in Dutch finally becoming consistent and accessible, mainly from the Flanders region of Belgium. This region nowadays speaks Flemish, a dialect of Dutch, which has some differences between standard Dutch from the Netherlands (Willemyns 2013). During this initial period of Dutch growth however, there was a great deal of influence from the French language in Dutch speaking territories, which came from the upper class (Marynissen and Janssens 2013). This was particularly prevalent in Belgium, where the French speaking Burgundians decided to use Brussels as their main base of residence, influencing nearby places in the rest of Flanders and the Southern Netherlands (Marynissen and Janssens 2013). In fact, it was only from the 16th Century onwards that Dutch began a more official process of standardisation, with various cities in the Netherlands and Belgium adopting a standard Dutch form. This also coincided with rich cultural development in the Netherlands, with esteemed artists, writers and scientists all bringing Dutch into the mainstream (Marynissen and Janssens 2013). By the end of the 18th Century, Dutch had become accessible to all classes and demographics in the Netherlands and Belgium, as well as spreading overseas to Suriname and the Dutch Antilles. On top of this development, came a second attempt at standardisation, in the form of ABN (Algemeen Beschaafd Nederlands), or General Civilised Dutch (Marynissen and Janssens 2013). In fact, in comparison to English and German, Dutch has far fewer dialects and varieties, as well as fewer speakers, with around 23 million natives worldwide and an extra 5 million speaking it as a second language. Amongst Dutch natives from the Netherlands and Belgium, it was found that both nations preferred their own variety, however, Netherlands Dutch was linked more with status and power, while Belgian Dutch seemed to be friendlier and beautiful (Lybaert and
21
Delarue 2017). Looking at wider attitudes to Dutch, research by Van Bezooijen (2002), found that Dutch nationals tended to view their own language and German as unpleasant, while languages such as Italian, Spanish and French are regarded as beautiful. Apart from Dutch’s other varieties, German is perhaps its closest relative with their being many similarities lexically, although there are also many differences outlined in Table 8.
Table 8. Lexical differences between Dutch and German.
English Carrot Bike Listen But
Dutch Wortel Fiet Luisteren Maar
German Karotte Fahrrad Zuhören Aber
As well as the lexical differences outlined above, there are also some great differences in phonology, such as the ‘sch’ sound outlined by Van Koppen (2014) in section 2.1. Considering their closeness, it could be that Dutch and German end with similar ratings, however Gooskens and Van Heuven (2019) found that Danes have very negative opinions of the Dutch language, scoring it well below 2, while Swedes also only scored it just above 2. Considering the low opinions of other Germanic language speakers, it could be that Romance language speakers hold even more negative viewpoints. Alternatively, there could be a very low opinion of Dutch in Scandinavia, with opinions elsewhere being more positive.
Danish and Swedish
The next two languages, Danish and Swedish, share a great deal of overlap so will be discussed at the same time. Starting with Danish, this comes from the East Scandinavian branch of the Northern Germanic languages and like the other Scandinavian languages, was originally written with a runic script (Encyclopaedia Britannica 2013). Danish and Swedish remained incredibly similar up until around the 11th century when they began to diverge (Lentzner 1895). Following this, the first Danish law text was written in the 12th century and by the end of the 14th, Danish overtook Latin as the main language of administration in Denmark (Haberland 1994). Interestingly, for a language of only 5 million speakers, it is rather wide spread, being spoken by minorities in Greenland, the Faroe Islands and Germany, while it is also mutually intelligible with Norwegian and Swedish, especially when written (Haberland 1994). In fact, the Southernmost region of Sweden, known as Scania, used to be a Danish speaking area (Haberland 1994). Resultantly, the Swedish now spoken in this area is much closer to Danish
22
than in other areas of the country. Indeed, despite receptive multilingualism usually being used between the nations, some debate still exists over Danish and Swedish’s true mutual intelligibility, with language attitudes having a large role to play in this matter (Schuppert et al. 2016). Some lexical differences between Danish and Swedish are illustrated below:
Table 9. Lexical differences between Danish and Swedish.
English Girl Week Room Ankle
Danish Pige Uga Værelse Ankel
Swedish Flicka Vecka Rum Fotled
However, the major differences between the two languages occur phonologically. Research by Maurud (1976) found that Danes were able to understand about 60% of spoken Swedish, while Swedes were only able to understand around 48%, although this figure is likely to be higher in regions, such as Scania. One of the key differences comes in vowel pronunciation, with Haberland (1994) explaining that there are many vowel contrasts within the language that cannot even be transcribed by IPA without the aid of diacritics. An example of this was provided in section 2.1, in the form of the Stød (Haberland 1994). Summarising, although Danish and Swedish share significant similarities, there are certain features of Danish that create a divide between the two languages.
Some reference will now be made to Swedish, whose history overlaps significantly with that of Danish, starting out as a runic language of the East Scandinavian branch of the Northern Germanic family (Encyclopaedia Britannica 2013). Contrary to Danish, Swedish did not hold a huge amount of power in its early and middle history, with Danish being further ahead in its developmental stage and holding political and economic power in much of Scandinavia. However, in the 16th Century, attempts to lessen the impact of Danish and bring Swedish to the fore as its own independent language were put in place, becoming standardised soon after and even beginning to spread at the expense of Danish (Encyclopaedia Britannica 2013). In the present day, there are over 10 million Swedish speakers across the world, with it having recognised official status in neighbouring Finland as well.
Unfortunately, there is very little research into attitudes of Swedish and Danish outside of the two countries, with research indicating that Swedish is the more beautiful of the two languages. In support of this, Schuppert et al. (2016) found that certain factors, such as speed and pitch of
23
voice can indicate why Swedish is more beautiful. For example, Schuppert et al. (2015) suggested that Swedish being a pitch accent language, results in more favourable opinions in comparison with a stress-language, such as Danish. Furthermore, Schuppert et al. (2016) also report on the fact that Danish being a faster language is also a reason behind its more negative attitudes as it causes even more intelligibility problems with Swedish. With these factors in mind, it seems that Danish may have a more monotonous pitch in comparison to Swedish, while the speed at which they talk makes comprehension difficult, especially when considering the bizarre nature of some of the languages sounds, for example, the Stød.
Referring to Gooskens and Van Heuven’s (2019) research, Danes gave Swedish a 4 out of 5 on the beauty scale, whereas Swedes gave Danish just over 2.5. This also took into consideration other Germanic languages, all of which scored Swedish higher than Danish. Having highlighted the relationship between the two languages, as well as some reasons as to why Swedish may be more beautiful, it will be very interesting to consider how attitude scores amongst non-Swedish, non-Danish and non-Germanic speaking people will record.
3. Research Questions
Various research questions can now be proposed in relation to some of the ideas and hypotheses that were discussed in the background section. These are as follows:
1. Which language and language family is the most beautiful?
In the background section, each languages development, history and global impact has been referenced, as well as recent research from Gooskens and Van Heuven (2019), to understand how present-day attitudes towards these languages might be formulated. This question will act as the studies focal-point, presenting findings on how each language and language family is rated, using the background research into each language, as well as Giles et al. (1974) and Trudgill and Giles’ (1978) hypotheses to explain why this may be.
2. Do attitude ratings differ pre and post listening to the speaker’s recordings?
In line with the hypotheses of Giles et al. (1974) and Trudgill and Giles (1978), as well as various research into explicit and implicit attitudes (Beckler et al. 2006; Greenwald and Banaji
24
1995; Wittenbrink and Schwarz 2007), pre ratings will be understood more as participants’ explicit ratings of each language that are influenced by imposed norms and social connotations, while the post ratings will be seen as more implicit attitudes towards each language, guided less by stereotypes or explicit attitudes and more by how the language sounds when spoken by a native.
3. Do attitude ratings differ for female and male speakers?
Some ideas on what makes a voice attractive has been outlined by Trouvain and Zimmerer (2017). Considering these, attitude scores given for each gender will be assessed to see if there is a difference and explore whether a certain genders voice has more attractive features than the other.
4. Do male and female participants give different scores to male and female speakers?
While question 3 will look at if there is a preference for a certain genders voice, this question will aim to assess the impact the gender of the participants has on this. Research from Trouvain and Zimmerer (2017) has suggested that there may be a preference amongst males and females for voices of the opposite gender and this will be explored.
5. Does a participant’s linguistic background impact the attitude ratings they give to other
languages?
In line with the imposed norm hypothesis (Giles et al. 1974), and the social connotations hypothesis (Trudgill and Giles 1978), this question will investigate whether speakers of certain languages prefer the varieties of their own language family, or of another. This will also test Van Bezooijen’s (1994) hypothesis of intelligibility leading to higher attitudes.
6. Does a languages identifiability lead to different attitudes towards it?
In line with research from Deutscher (2010), it will be interesting to see if there is a disparity between languages that are easily identified receiving high ratings and languages that are more difficult to identify receiving low ones.
25 3.1 Hypotheses
Referring to the background section and the literature that was discussed, several hypotheses, related to each research question can also be proposed:
1. Romance languages will be rated higher than Germanic languages. A Romance language
will finish as the most positively rated language.
This has been hypothesised in line with various research that has suggested that Romance languages closer relationship to Latin leads them to having higher historical prestige and therefore higher attitude ratings (Penny 2002; Hanners 1999). Additionally, attitudes towards French, Spanish and Italian seem very positive (Gooskens and Van Heuven 2019; Nadeau and Barlow 2008; Giles and Niedzielski 1998), although less clear towards Portuguese and Romanian. In comparison, attitudes towards the Germanic languages are more varied, with high opinions being hypothesised for English (Jenkins 2009; Rindal 2015), but much lower opinions potentially coming for German, Dutch and Danish (Haberland 1994; Schuppert et al. 2016; Batram 2010).
2. Ratings for the Germanic languages will go up after listening to the speaker’s recordings,
while attitudes towards Romance languages will remain constant.
This is hypothesised, as many of the negative stereotypes towards certain Germanic languages tend to index the negative historical development of these languages (Dekker et al. 1998; Henriksen and Van der Auwera 1994), which may not be as prevalent in modern-day society. As these stereotypes are often referenced when giving explicit opinions of a language, the more implicit opinions that will be given post-recording may be more positive as they will be influenced less by stereotypes and more by their opinion of the sound of the language and the speaker. Romance languages are expected to remain constant in their ratings, as opinions and stereotypes of the languages in this family tend to be positive (Gooskens and Van Heuven 2019), while the languages seem to have fewer identifiable ‘ugly’ features than the Germanic languages (Penny 2002; Giles and Niedzielski 1998).
26
Trouvain and Zimmerer (2017) outlined several attractive voice traits that can affect attitudes towards a speaker and considering Lakoff’s (1975) idea that females are more careful with their speech, it is expected that females may have more attractive features in their speech and voice than males, who may be less concerned with sounding attractive.
4. Males will give higher scores to female voices and vice versa.
This will use Trouvain and Zimmerer’s (2017) idea that hetero-sexual adults prefer voices of the opposite gender and explore this in a more general sense.
5. Participants who speak Germanic languages will rate Germanic languages higher than
the other group of speakers.
While there is some evidence to suggest that speakers of Germanic languages do not favour their own languages, research by Van Bezooijen (1994) suggests that listeners prefer languages that are close to their own. Based off this, as well as already negative attitudes towards certain Germanic languages, it is believed that Germanic speakers will rate their own languages more pleasantly than other speakers who are more unfamiliar with them.
6. Languages that are harder to identify will receive lower attitude ratings.
This is based off Deutscher’s (2010) belief that a language being more familiar amongst a general population, will lead to it having higher opinions, while less well-known languages with rarer sounds might receive lower attitude ratings.
4. Materials
The materials required for the survey were 20 recordings of each language being spoken by a native speaker of that language. Figure 2 illustrates where each speaker that provided a recording is from, male speakers in blue, female speakers in red:
27
Figure 2. Map of Speakers' birthplaces.
The reason for there being 20 recordings was so that a female and male speaker could represent each language. Friends of mine, natives in each language, were contacted with information about the recording and the survey. Here, it is important to note that while the speakers were all natives, they were not all necessarily standard speakers of the language, coming from a variety of different areas, with different accents and dialects. Once they consented to participating, they were sent a short text to record and send back to me. Prior to this, they were given instructions to speak clearly so that the clips would be audible and coherent. For certain languages, I did not have a contact who spoke it to a native level, so certain speakers were recruited through the contacts of friends. As this was a study into European languages, the speakers also had to be from a European country where it was an official language. For example, only Portuguese speakers from Portugal were considered, not Brazilians.
28
The text that featured as the recording in the survey was taken from the European Union Languages webpage, which had a text translated into the 24 official languages of the European Union (https://europa.eu/european-union/about-eu/figures/administration_en). As this study tested only Germanic and Romance languages, these were the only ones that were selected from the 24 available. The last paragraph of this text was chosen, as it was around two sentences long, and would not require too much recording or listening time for speakers and participants. The fact that the text was written in the different languages was highly beneficial because no texts had to be translated, as they had already been written and styled by native speakers of that language. The English version of the text is as follows:
“In the increasingly interdependent world of the 21st Century, it will be even more necessary for every European citizen to co-operate with people from other countries, in a spirit of curiosity, tolerance and solidarity.”
Upon receiving the clips, they were edited using an Audio-cutter to ensure that they started and finished with no pauses. Clip length varied, presumably due to factors, such as word count and length, as well as speech rate. As an example, the longest recording was made by a Dutch speaker, which lasted 18 seconds, while the shortest was made by a Swedish speaker, lasting just 9 seconds. However, looking at the two texts, it is clear why one would take longer than the other:
Dutch
“In deze eenentwintigste eeuw leven we in een wereld waarin we steeds afhankelijker van elkaar worden. Het is dan ook noodzakelijker dan ooit tevoren dat iedere Europeaan zich openstelt en in een geest van tolerantie en solidariteit samenwerkt met mensen van andere landen.”
Swedish
“Under 2000-talet kommer det att bli allt viktigare för varje europeisk medborgare att kunna samarbeta med folk från andra länder i en anda av nyfikenhet, tolerans och solidaritet.”
