Constructed Language: An Analysis of the Phonemic
Sounds Influenced by Historical Stereotyping
Author: Ashlie Devenney
MS Date: 05-30-2017
FL Date: 07-01-2017
FL Number: FL-000046-00
Citation: Devenney, Ashlie. 2017. » Constructed
Language: An Analysis of the Phonemic
Sounds Influenced by Historical
Stereotyping.» FL-000046-00, Fiat Lingua,
Copyright: © 2017 Ashlie Devenney. This work is licensed
under a Creative Commons Attribution-
NonCommercial-NoDerivs 3.0 Unported License.
http://creativecommons.org/licenses/by-nc-nd/3.0/
Fiat Lingua is produced and maintained by the Language Creation Society (LCS). For more information
about the LCS, visit http://www.conlang.org/
.
1
Constructed Language: An Analysis of the Phonemic Sounds Influenced by Historical
Stereotyping
AP Research
10 April 2017
(Word Count: 4010)
2
Table of Contents
Abstract……………………………………………………………………………………………………………….4
Introduction………………………………………………………………………………..5
Review of Literature………………………………………………………………………5
Constructed Languages……………………………………………………………5
Perception…………………………………………………………………………6
Media………………………………………………………………………………7
Construction of Constructed Languages…………………………………………………………9
Methods……………………………………………………………………………………9
Survey……..………..……………………………………,…………………..…..9
Statistical Analysis……..…………………………………………………………..9
Transcription…………………………………………………………..….………11
Results and Discussion……………………………………………………………………12
Black Speech Phonology Comparison………………………………………..….13
Sindarin Phonology Comparison…………………………………………………15
Klingon Phonology Comparison……..………………………………………….15
Shiväisith…………………………………………………………………………15
Conclusion and Future Discussion……………………………………………………….16
Conclusion……………………………………………………………………….16
Significance………………………………………………………………………16
Limitations……………………………………………………………………….17
Delimitations……….………………………………………….…………………17
Future Directions…….……………….…………………………………………..18
Acknowledgements………….…………..……………………………………………..19
Appendix……………………………………….………………………………………20
Works Cited………………………….………………………………………………….21
3
ABSTRACT
4
The perception of constructed languages in film is not a topic that has been researched
extensively in the past due to the scrutiny concerning the field of constructed languages as a
valid field of study. An understanding of how humankind perceives constructed languages is
vital in our understanding of how natural languages are perceived. The purpose of this research is
to examine how the base phonemic sounds of a language (particularly constructed languages)
affect how the listener hears and perceives a constructed language as well as how and why this
perception is constructed. This study is done through a survey consisting of several languages
wherein the participant rates the languages on certain qualities which establish how the
participant feels towards the language. The research finds that a historical relationship between
the beginnings of language construction and the listener’s perception of that language,
discovered through an analysis of the phonemic sounds, exists in both constructed and natural
languages. This finding will help those who create constructed languages determine what sounds
need to consistently occur for their language to be perceived according to intention.
5
INTRODUCTION
Constructed languages are often used in science fiction films to highlight the personalities of the
characters who speak them. However, how our perception of these characters is influenced by
their respective constructed language has yet to have been explored. To enhance the
understanding of natural human language and how we perceive it, we can analyze how
perceptions of constructed languages are influenced by their phonemic idiosyncrasies which may
inform writers in the creation of a constructed language that best represents a particular character
to an audience. Clive Upton conducted a study similar to this one wherein he determined that
accents are discriminated against based on aesthetic grounds (Upton). In this research, I focus on
how the phonemic sounds of a constructed language affect how the audience perceives the
language, and how that relates to historical circumstances. I attempt to advance the idea that
accents are discriminated against based on aesthetic grounds, by concluding that individual
phonemic sounds affect the perception of language, specifically those languages which do not
occur naturally, through a survey.
REVIEW OF LITERATURE
Constructed Languages
Even though the popularity of constructed languages has skyrocketed as science fiction
films have become increasingly popular within the last decade, the study of constructed
languages is not as new of a field as some may believe. There is a rich history of the creation of
constructed languages dating back to 600 BCE (Eco), including the legendary origins of the Irish
language. The legend tells of a man named Fenius Farsaid, who, along with 72 fellow scholars,
traveled from Scythia to the Plain of Shinar to study the languages that were confused at
Nimrod’s tower (commonly known as the Tower of Babel) (Nyland). After discovering that
6
these languages were spread throughout the world, Farsaid sent out the scholars to study the
languages which were dispersed (Nyland). Ten years later, the scholars returned with their
findings and Farsaid, with the help of these scholars, took the most favored parts of each of the
languages and created what he referred to as “in Berla tobaide” or “the selected language,” which
he then began calling Goidelc (Nyland). Though the creation of this language is only considered
a legend today, the idea of a constructed language was in place. The mythical aspect to the field
contributes to the skeptical criticism regarding the field and those who study it.
Often times, the field is looked down upon as it does not concern the study of native
languages. In the eyes of some, linguists should focus on saving the languages that are currently
endangered instead of making up new languages. However, Marc de Oostendorp, a Dutch
linguist, was critical of this viewpoint which is evident in his writing, “Constructed Languages
and Linguistic Theory” (Oostendorp). In his piece, Oostendorp argues that constructed or
planned languages “have been made up in slightly different ways” and “serve different goals”
and are therefore valuable in ways that differ from the value of natural languages. (Oostendorp).
Though it is important to protect the languages which currently exist, it is beneficial for linguists
to study and construct languages in order to further the understanding and appreciation of how
language as a whole, functions. Constructed languages are imperative in our understanding of
natural languages and how we perceive them. It is due to the debate over the value in studying
constructed languages that we find areas that are muddled, where only lack of research is to
blame.
Perception
Having established the importance of constructed languages in the linguistic field, we can
now turn to the stereotypes we have and judgements we make based on that arise from perceived
7
linguistic features and whether those judgements were considered during the creation of a
language, particularly in science fiction films. As Dr. Lisa Davidson, a professor of linguistics at
New York University states, people tend to refer to languages with sounds primarily made in the
back of the vocal tract as “harsh-sounding” (Cronim). However, Davidson also stated that the
perspective of the listener will play a role in how the language sounds and could make the
language sound harsher than a person with an unbiased perspective (Cronim). For example, the
velar fricative phonemic sound is commonly found in the German language and may not be
considered a harsh sound to a native speaker, but to an outsider, the sound might be classified as
such. A professor of German Linguistics at the University of California made the point that
negative stereotypes of a speaker will have an impact on how the language is perceived by the
listener (Cronim). A piece by Clive Upton, a professor of English at the University of Leeds in
England, came to a similar conclusion when he, in The Oxford History of English, gives attention
to the fact that certain accents are viewed unfavorably based on aesthetic grounds (Upton). Given
that accents are only differences in pronunciation, this research takes this idea one step further by
correlating the idea of unfavorable judgement based on accents to unfavorable judgement of
entire languages due to the phonetic sounds.
Media
Movies have a large impact on their viewers, in more ways than one: for example,
movies can shape political views, moral beliefs, and important for this research, the way people
subconsciously perceive languages different from their own. The movies that I will incorporate
into my research are The Lord of the Rings franchise, Thor: A Dark World, and Star Trek: Into
Darkness.
8
The movie The Hobbit: An Unexpected Journey was chosen from The Lord of the Rings
franchise. The movie is set prior to the original Lord of the Rings series, wherein the Hobbit,
Bilbo Baggins, first obtains the ring which causes the entirety of the problems in the rest of the
series (Jackson). The reason this movie is incorporated into this study is because this is one of
the movies where the audience is presented with the native Orc language (Jackson), known as
Black Speech. In this same movie, the audience is also presented with the language of the Elves,
specifically the Sindarin dialect, which will allow for a comparison the two languages with fewer
variables between them (Jackson). In the franchise, the Orc race is seen as the enemy while the
Elfin race is portrayed in a positive light and is who the main characters rely on for assistance
several times throughout the storyline (Jackson). In a book written by Michael Adams entitled
From Elvish to Klingon: Exploring Invented Languages, Adams makes the claim that Tolkien
(creator of the Lord of the Rings series) created the Elvish language to be “specifically pleasant”
(Adams). This claim relates directly to the research question which is whether or not the general
sound and formation of the language, on a historical premise, has anything to do with the image
the author wanted to present for that particular race The next movie used in this research is Thor:
A Dark World which is about the popular Marvel superhero Thor, a god of the Asgard realm,
who clashes with the Dark Elves, a race that wants to unleash an all-powerful weapon upon all
the realms in this fictional universe (Taylor). This movie was chosen because the Dark Elves
have a very guttural-sounding language, and its use in this research strengthens the hypothesis
that languages with exceedingly guttural sounds will cause the listener to perceive the language
more negatively than they would perceive a language without this linguistic feature. The last
movie that will be included in this study is Star Trek: Into Darkness. The movie details Captain
Kirk’s problematic mission to capture a powerful weapon of mass destruction
9
(Abrams). This movie was chosen because of its popular constructed language, Klingon. Klingon
is an antagonist language in the series, but is well-known as a constructed language. Though as
little bias as possible is desired, consideration will be given to Klingon due to its prominence
among lovers of science fiction, and variances between the perception of Klingon and other
languages may be attributed to its popularity.
Construction of Constructed Languages
David J. Peterson, renowned conlanger, states in The Art of Language Invention that a
conlanger can use “the expectations of their users/hearers to create a phonaesthetic effect”
(Peterson). Conlanger Mark Rosenfelder agrees, pointing out that “authors…make use of what
we might call phonetic stereotypes” (Rosenfelder). However, in The Art of Language Invention,
Peterson discusses the importance of creating a language based on the speaker of that language
(Peterson), while Rosenfelder emphasizes using stereotypes primarily for the listener to perceive
the speaker in a particular way (Rosenfelder). The understanding that in a fictional realm, the
speakers of a conlang have several traits, is crucial when creating a language. Hence, it is
important to hinge one’s language on the speakers of that language and their qualities,
personalities, etc. The indicated discussion establishes a foundation for this research which aims
to determine how a conlanger can use the “phonaesthetic affect” (Peterson) to give depth to an
alien speaker.
METHODS
Survey
To begin exploring the question of the connection between linguistic features and listener
perception, I conduct a survey wherein participants are provided with an audio clip of the
languages used in this study, and then rate them on a scale from one to seven on various qualities.
10
A survey is essential in answering the research question thoroughly, given that using only my
perception of a language would not provide data that could be applied, even to some extent, to
the general population. The four qualities that participants rate the languages on are kindness,
attractiveness, intelligence, and honesty. These qualities were selected because all are quick
judgements an individual tends to make about the people they encounter, generally based on
their speech and body language. These qualities aid in determining whether or not the participant
judges a language, (and thus its speakers), favorably. Audio clips were chosen instead of video
clips in order to curtail the likelihood that the participants might recognize the language or the
movie the language is from. To eliminate as much bias as possible, the participants were asked
whether they recognize the language, and if they do, to what extent that recognition affects their
ratings.
Statistical Analysis
After the survey was conducted, the instances wherein the participant spoke/recognized
the language were removed in order to eliminate the bias that comes from knowing what
language the clip shows before rating it. To determine the statistical significance of the
relationship between constructed languages and the natural languages that were rated in the
survey, this research made use of the Chi-Square statistics test. The Chi-Square test is used to
assist a researcher in determining if any significant relationship exists between any two
categorical variables. The goal in performing the test was to find the amount of similarity that
exists between rating the qualities of the constructed language versus the natural language. The
null hypothesis for this research is that there exists no difference in the way a participant rates a
constructed language versus a natural language. The alternative hypothesis states that there exists
a difference in the way a participant rates a constructed language versus a natural language. This
11
test informs the researcher of which constructed languages and which natural languages the
population considers similar, based on survey results.
Transcription
After performing the Chi-Square test, this research requires a transcription of the
languages into the linguistic International Phonetic Alphabet, an alphabet used to document
languages based on the sounds that are used when speaking them. Based on this transcription, I
found patterns relating the survey participants’ ratings to certain sounds and then compare those
sounds to the languages of foreign enemies at the time that the constructed language was
invented. In order to find which phonemic sounds are particularly characteristic of a language, I
determined the average amount of occurrences a given sound has in the language, and then
separated the sounds which occurred more than the average, and used those to clearly define a
relationship between languages. From there, a historical analysis predicated on stereotypes was
made by comparing the characteristic sounds of the constructed language and the natural
language that were rated similarly in more than one category. If the two languages were rated
similarly in only one category, they were excluded from the Chi-Square test analysis. This is
because a situation such as that could be attributed to coincidence.
12
RESULTS AND DISCUSSION
Languages Compared
Friendliness Intelligence Honesty Attractiveness
Sindarin vs. German
Sindarin vs. Korean
Sindarin vs. Russian
–
–
–
Klingon vs. Korean
4.715
Klingon vs. German
2.337
–
–
.640
4.248
2.949
2.633
–
–
–
–
–
–
–
–
4.283
Black Speech vs. Russian
2.875
4.124
1.025
–
Black Speech vs. Korean
2.170
–
3.212
4.620
Black Speech vs. German
–
4.692
4.066
1.975
Figure 1: Lowest Chi-Squared Values for Language Comparisons
The comparisons that resulted in the lowest Chi-Squared values are shown above in Figure 1. All
values less than five are included in the chart. The smaller the Chi-Square value, the more similar
the two languages were rated. When the test was run, there were no comparisons for Shiväisith
which resulted in a Chi-Squared number of five or less, therefore it is excluded from the table.
This is also the case with the comparison of Klingon and Russian. A complete table of all the
Chi-Squared values can be found in the Appendix. The values shown in Figure 1 are relevant to
this study because this research requires an analysis of the most-similar ratings in order to judge
similarity of perception. From the table, it is clear that the comparisons of Black Speech with
each of the natural languages are the most similar, given that three of the four traits have a value
13
of less than five.
Black Speech Phonology Comparison
The phonology of Black Speech will now be discussed in order to highlight a more direct
comparison between natural languages and constructed languages. Black Speech will be
compared with all three of the natural languages because it was rated similarly in three of the
four categories for each of these natural languages. First, a comparison between Black Speech
and German will be made. Black Speech was created in 1943 by J.R.R. Tolkien, the English
author of The Lord of the Rings book series (Tolkien Gateway), and was then expanded by
language creators for the purposes of the movie adaptations. Since Tolkien is English, it is more
relevant to use historical relationships with the United Kingdom as opposed to the United States,
though they do not differ greatly. Worthy of mention is the fact that this was during the World
War II era, which lasted from 1939 until 1945 (Networks). The United Kingdom was involved in
this conflict with Germany, along with other countries fighting against Hitler’s regime. From this
conflict, we have developed several stereotypes of what qualities constitute a German.
Stereotypes such as the evil German scientist and even more broad concepts such as the idea that
“Germans have no sense of humor” or the ever-so-popular, “Germans are always angry.” None
of these stereotypes have anything to do with Germans themselves, and everything to do with
how society perceives them due to their role in historical conflicts. Once we begin to perceive a
group of people a certain way, everything about them becomes attached to that perception,
including their language. Based on the transcription of the dialogue presented in the survey, it
was found that the consonant phonemes common in the language are /k/, /z/, /dh/, /n/, /g/, and /r/
(note that these are all phonemes in the Black Speech language, however the phoneme /dh/ does
14
not occur as a single phoneme in English). The transcription of the German dialogue used in the
survey into IPA shows that the consonant phonemes that are characteristic of the language are /z/,
/d/ (a close phoneme to the /dh/ in Black Speech), /n/, and /r/. During World War II, the opposing
side was headed by Germany meaning that tensions were at an all-time high. Therefore, it is
likely that Tolkien created this language, either subconsciously or otherwise, with influences
from the German language in order to further the antagonistic role that the Orcs in the series
represent in the series. Between Korean and Black Speech, there are no common characteristic
phonemes. These two languages were rated similarly, however; this study cannot claim that
Korean had any influence in the creation of Black Speech seeing that the two languages share
none of the characteristic phonemic sounds. This claim makes sense in the context of this study
because there were no significant conflicts between Korea and the United Kingdom until after
the creation of the language. In contrast, Black Speech and Russian share the characteristic
consonantal phonemes /r/, /z/, and /d/ (a close phoneme to the /dh/ in Black Speech). During
World War II (the time when Black Speech was created), Russia was an ally with the United
Kingdom, therefore citizens of the UK likely had a more positive perception of the Russian
people, thereby the language. Though the two languages share a few of these phonemes, this
research cannot make a claim as to whether or not Russian influenced the creation of Black
Speech. This is because one cannot assume he did not use it because they were an ally and he
was creating an antagonistic language, nor can one assume that he did use it for inspiration and
disregarded the perception of the Russian people and their language.
15
Sindarin Phonology Comparison
Turning toward the more protagonist language of Sindarin, the transcription of the
dialogue found that the characteristic consonantal phonemes are /θ/, /r/, /n/, /d/, and /l/. Sindarin
was created shortly after World War II by Tolkien for The Lord of the Rings book series, and
just like Black Speech, was expanded for the purposes of the films. (Tolkien Gateway). Though
Sindarin and Russian were rated similarly, they only share two characteristic consonantal
phonemic sounds which are /r/ and /d/. Therefore, a claim cannot be made that Tolkien used the
language of the Russians (who were allies of the United Kingdom during World War II) to mold
the protagonist Sindarin language, at least not to any significant extent.
Klingon Phonology Comparison
The characteristic consonantal phonemic sounds of Klingon are /d͡ ʒ/, /v/, /x/, /p/, /m/, and
/b/, /d/, /n/, and /r/. This language was rated similarly to German, therefore this study will
compare the phonology of the two languages and their relationship to one another. These two
languages share the characteristic consonantal phonemes /m/, /d/, /r/ and /n/. Klingon was created
to add dimension to Star Trek: The Motion Picture in 1979 (Klingon Language Institute).
Therefore, this language was created after the major conflict with Germany during World War II.
Given that the languages were created with several of the same characteristic consonantal
phonemic sounds, it can be determined that the creators of this language were using German for
inspiration during their creation process.
Shiväisith
Shiväisith did not present enough similarity in ratings to the natural languages to
establish a concrete relationship between the two. However, several factors go into the creation
of a language, and Shiväisith is no different. David Peterson, the conlanger (creator) of this
language, stated that he and the director wanted the language to be “reminiscent of, (but distinct
from) Finnish.” due to the comic and movie’s connection with Norse mythology (D. Peterson).
Factors such as these are also taken into consideration when constructing a language to add
16
dimension to the characters speaking it.
CONCLUSION AND FUTURE DISCUSSION
Conclusion
The phonemic sounds of a constructed language do have an impact on the way the
audience perceives the constructed language. Furthermore, it is evident, especially through an
analysis of the Black Speech language, that historical events involving natural languages
influence how a listener perceives a constructed language. It should be realized that historical
events and cultural stereotyping are not the only aspects that influence a conlanger’s construction
of a language. Only focusing on how the language will be perceived by an audience will not
construct a successful path to depicting a character’s true culture, which is a significant purpose
of constructing languages.
Significance
In context of the broader field of linguistics, this research aids in the understanding of
how the listener perceives language, which applies not only to constructed languages but natural
languages as well. Conlangers can go on to create a language based on specific phonemic sounds
and produce a result that will enhance the probability that the perception of the conlang-speaking
characters will meet the conlanger’s satisfaction. The conlanger can also attempt to change the
connotation of harsher sounds through use of these sounds for protagonist characters, thereby
eventually changing the perception of these sounds and moving past this surface-level
categorization. Through language, we see a division between “attractive” and “unattractive”
17
languages, which negatively enforces the need to categorize and stereotype. This classification of
languages, as shown in this research, seems to stem from historical backgrounds and not solely
the sounds of the language. This research should concern conlangers who want their language to
be perceived in a particular way; especially those working for the film industry who are asked to
create the language to fit a certain sound.
Limitations
The sound clips that were used in the survey were directly pulled from the movie itself,
therefore music and background noise likely played a role in how the listener perceived the
language. The survey included questions in order to limit bias as much as possible, however bias
could still be prominent if a participant recognized the language and failed to indicate this during
the survey. Any mistakes in the transcription of the dialogue or the counting of the results can be
attributed to mere human error. It should also be noted that the Chi-Squared test failed to present
values for a few of the comparisons, due to the fact that expected counts were not at the desired
minimum of five. Out of all of the conlangs, Black Speech was the language that yielded the
most results, therefore that is the language that was discussed to the greatest extent in this
research. This is a limitation in that the other constructed languages were unable to be analyzed
to as great an extent as Black Speech, however one can still see the relationship between
constructed languages and natural languages.
Delimitations
Taking into consideration the limitations of the study, we can now turn to the
delimitations. The survey received ninety-seven responses, a large enough sample size to get an
accurate picture of how an audience perceives these languages. Though the transcription of the
dialogues into the International Phonetic Alphabet may have errors, the phonology of the
18
languages was analyzed to the greatest possible extent. This helped to confirm or deny which
sounds were characteristic of each language. Though the Chi-Squared expected counts were not
where statisticians like them to be, the values that this study resulted in can still give the
researcher an idea of the relationship between a constructed and natural language.
Future Directions
For future study, it would be beneficial for the researcher to include more languages in
any research they conduct. This would allow for a more diverse survey; wherein more
historical/stereotypical connections could possibly be made. It should also be recommended that
any future surveyor use language clips of one individual speaking, as opposed to subjecting the
survey to the various tones and varying pronunciations of the actors who portray the alien
characters in film. Regarding the survey, it will benefit the researcher to randomize the clips of
sound so that the clips of language are not in the same order for all participants. This will
decrease the likelihood that the participant will become tired of listening to clips and rating them
as the survey progresses; therefore, causing participants to rate the languages less accurately than
the languages at the start of the survey. To make the process easier, any future researcher might
want to begin the survey by asking the participant what languages they speak and/or are familiar
with. This will eliminate confusion by removing responses where the participant recognizes the
language before conducting any further analysis. Lastly, the researcher would benefit from using
clips of natural languages from movies as opposed to introductory language videos. The
conlangs presented in the survey are more conversational whereas the natural languages
presented may be described as artificial. The natural languages attempt to help the audience learn
the language, therefore increasing the likelihood that the language sounds pleasing to the ear.
19
ACKNOWLEDGEMENTS
I would like to acknowledge and thank Dr. Jessie Sams of Stephen F. Austin for her assistance
throughout the research process. I would also like extend a special thanks to David Peterson for
his time, technical contributions and overall advice.
Appendix
20
Languages Compared
Friendliness Intelligence Honesty Attractiveness
Shiväisith vs. German
N/A
9.688
11.301
N/A
Shiväisith vs. Korean
N/A
5.532
20.775
31.451
Shiväisith vs. Russian
N/A
16.02
5.067
5.139
Sindarin vs. German
11.292
5.568
.640
N/A
Sindarin vs. Korean
5.217
9.661
4.248
30.825
Sindarin vs. Russian
N/A
2.949
2.633
6.335
Klingon vs. German
2.337
8.515
6.844
4.283
Klingon vs. Korean
4.715
7.529
10.628
N/A
Klingon vs. Russian
9.661
10.465
7.736
8.437
Black Speech vs. German
N/A
2.283
4.066
1.975
Black Speech vs. Korean
2.170
4.692
3.212
4.620
Black Speech vs. Russian
2.875
4.124
1.025
18.159
Figure 1. Table of all the Chi-Squared Values for the Language Comparisons
21
Adams, Michael. From Elvish to Klingon: Exploring Invented Languages. New York: Oxford
Works Cited
University Press, 2011. Web.
Conlangs. Dir. Ben Wood. Perf. Grant Goodall, et al. 2015. Web.
Cronim, Melissa. German Sounds Harsher than Other Languages and Here’s Why (Video). 3
August 2013. Web. 23 January 2017.
Hinton, Leanne. The Green Book of Language Revitalization in Practice. San Diego: Academic
Press, 2013. Web. 22 3 2017.
Klingon Language Institute. 28 December 2016. Web. 8 April 2017.
Networks, AE Television. History Channel. 2017. Web. 7 April 2017.
Nyland, Edo. Linguistic Arcaheology. FriesenPress, 2016. Web. 12 January 2017.
Oostendorp, Marc de. «Constructed Language and Linguistic Theory.» Schubert, Klaus. Planned
Language: From Plan to Reality. 2000. Web.
Peterson, David J. The Art of Language Invention. New York: Penguin Books, 2015. Print. 21
March 2017.
Peterson, David. Research Involving Constructed Langauges. 13 January 2017. Web.
22
Rosenfelder, Mark. The Language Construction Kit. 1996. Web. 1 April 2017.
Star Trek: Into Darkness. Dir. J.J Abrams. Perf. Chris Pine, et al. 2013. Web.
The Hobbit: An Unexpected Journey. Dir. Peter Jackson. Perf. Martin Freeman and Ian
McKellen. 2012. DVD.
Thor: The Dark World. Dir. Alan Taylor. Perf. Chris Hemsworth and Natalie Portman. 2013.
Web.
Tolkien Gateway. 13 December 2015. Web. 7 April 2017.
Eco, Umberto. The Search for the Perfect Language. Vol. 1. Wiley, 1995. 1 vols. Web. 12
January 2017.
Upton, Clive. «Modern Regional English in the British Isles.» The Oxford History of English. Ed.
Lisa Mugglestone. New York: Oxford University Press, 2006. 305-333. Web.
