A Vision of Translation as Play and Game

The article discusses application of game theory to the study of language interaction and specifically to the study of translation. The terms that denote game and play in different languages are presented, as well as a number of conclusions with regard to representation of the translation process, based on the observation of students’ activities, performed online, as well as on the results of the survey among both translators and nontranslators. We have taken up one of the most relevant issues in modern translation studies—definition of translation as process. We have also analysed language means used to build TRANSLATION GAME mental space, as well as key aspects of activities performed by the translator, who is the key figure of the process. Special attention is paid to risk, which is the most important element of the translator’s episteme and the determining factor for the intentional horizon of the translator. Looking for the variant, solution, which would guarantee the best possible effect on the receivers, the translator sometimes has to take risks and even sacrifice parts to keep the integrity of the original message. Semiotics, game theory and decision-making theory together seem to offer the best instruments to analyse the issues of translator as homo ludens.


Introduction
Translation studies as an academic discipline is related to the study of the theory and phenomena of translation.By its nature it is multilingual and also interdisciplinary, encompassing any language combinations, various branches of linguistics, comparative literature, communication studies, philosophy and many more.Over time the interdisciplinarity of the subject has become more evident and recent developments have seen increased specialization and continued importation of theories and models from other disciplines, among them being also mathematics and game theory.
The focal topic of this paper is discussion of the identity of translator as key figure of translation game.The satellite topics are concerned with the process of translation and its elements in the light of game and play, as well as the way the process of translation is perceived both inside the translators' community and outside it.
Translation studies have expanded explosively in recent years.Time has come to go beyond restricted vocabulary and syntax and sharply restricted text types in building the model of the process of translation.We do not dare say that turning to the instruments of game theory will solve the tantalizing problem of understanding the essence of the process, but it will certainly help educate better translation experts.We have taken up one of the most relevant issues in modern translation studies-definition of translation as process.We have also analysed language means used to build TRANSLATION GAME mental space, as well as key aspects of activities performed by the translator, who is the key figure of the process.Special attention is paid to risk, which is the most important element of the translator's episteme and the determining factor for the intentional horizon of the translator.Looking for the variant, solution, which would guarantee the best possible effect on the receivers, the translator sometimes has to take risks and even sacrifice parts to keep the integrity of the original message.Semiotics, game theory and decisionmaking theory together seem to offer the best instruments to analyse the issues of translator as homo ludens.

Translation in the Light of Game and Play
Equivalence was a key word in the linguistics-based translation theories of the 1960s and 1970s, although its basic mode of thought may be traced back to Cicero and later to the Renaissance theories that began to presuppose languages of equal status.A close look reveals that some theories assume pre-existing equivalents and are thus concerned with a search for "natural" equivalence.Other theories allow that translators actively create equivalents, being concerned with "directional" equivalence.These two approaches are often mingled, giving rise to many misunderstandings and unfair criticisms of the underlying concept.The historical undoing of the equivalence paradigm came when the directional use of the term allowed that equivalence didn't need to be a belief or expectation at the moment of reception, which need not be substantiated on the level of linguistic forms.At the same time, source texts became less stable and languages have been returning to more visibly hierarchical relations, further undermining the concept (for more details see for example A. Pym (2009)).
A lively interdisciplinary community has emerged in recent years, which uses game theoretic techniques to study genuinely linguistic problems.When trying to analyse or define the concept of game and play one must always bear in mind that the idea as we know it is defined and perhaps limited by the words we use for it.Words and ideas, as J. Huizinga puts it, are not born of scientific or logical thinking, but of creative language (Huizinga, 1949).Nobody will expect that every language, in forming its idea of play and game and expression for it, could have come across the same idea or would have found a single word for it, just as there is one definite word for mother or father.However, the matter is certainly not as simple as that.

Huizinga defines the concept of play as follows:
play is a voluntary activity or occupation executed within certain fixed limits of time and place, according to rules freely accepted but absolutely binding, having its aim in itself and accompanied by a feeling of tension, joy and the consciousness that it is "different" from "ordinary life" (Huizinga, 1949).
Huizinga believes that thus defined, the concept is capable to embrace everything that is called play in animals, children and adults and various kinds of games.He ventures to call the category play one of the most fundamental categories of life.
It is quite remarkable that the word ludus (from Latingame, play, sport) has not only not passed into the Romance languages, but has left hardly any traces there.Ludus has been supplanted by a derivative of jocus, which extended its specific sense of joking and jesting to game and play in general.Thus, Romanian has joc, joaca, a juca, French has jeu, jouer, Italian gioco, giocare, Spanish juego, jugar.It is only in adjectives (ludic, ludique, ludico) that we see traces of ludus.The issue of whether the disappearance of ludus and ludere is due to phonetic or semantic reasons is beyond the scope of the present article.
The Germanic languages seem to have no common word for play or game.J. Huizinga believes that this may be explained by the fact that in the hypothetical archaic Germanic period play had not yet been conceived as a general idea.As soon as each individual branch of the Germanic languages came up with a word for play, these words all developed semantically in exactly the same way, or rather this extensive and seemingly heterogeneous group of ideas was understood under the heading of play (Huizinga, 1949).
The English words play, to play are very remarkable from the semantic point of view.Etymologically the word seems to come from the Anglo-Saxon plega, plegan, meaning primarily play or to play, but also rapid movement, a gesture, a grasp of the hand, playing a musical instrument and other activities. 1 Modern English still preserves much of this wider significance. 2he word game, even though often used synonymously with play, also has the meaning of procedure or strategy used to gain an end, a competition. 3ame theory proposes to study the behaviour of two or more people with conflicting interests, as in a competition.The theory was initially formulated by John von Neumann and Oskar Morgenstern in the work Theory of Games and Economic Behaviour.The simplest model of game is the two-person, zero-sum game with perfect information and optimal strategy.In this kind of games there are two players, the game ends after a certain number of moves have been made, there is always a winner and a loser and there is a strategy that allows the player who makes the first move to win irrespective of the moves made by the other player.Formal game theory also operates on the assumption that players act rationally.
Most players and social actors aim for optimal strategies on the basis of imperfect information.Players arrive at this optimal strategy by way of a pay-off matrix, a formal device that lists the alternatives and strategies available to players and allows them to evaluate outcomes so that they can choose the optimal strategy.Far from being limited to games in a limited sense of the word, game theory is the mathematical study of rational social interaction and, as such, it is reasonable to expect it to be able to shed light on language use as well.Perhaps more than anything, it promises to have the potential to explain why communication works the way it does.For if we could show that people's linguistic behaviour conforms with what it would be rational for them to do, then this would have substantial explanatory value.
The question of optimization appears to be central to the theory and practice of translation.We shall define translation similarly to K. Reiss and H. Vermeer.Translation to us is production of both written and oral discourse intended to render a message expressed in one language into another language.Translation to us is an infinite game, which means that rules of it may change at any time (for details about finite and infinite games of translation see J. P. Carse (1986)).The dramatic paradox of the infinite game is in the fact that you can only have something if you pass that something to others.So far we haven't managed to coin a better term for this activity.The game of translation does not fully reflect the content of this activity, unfortunately.It is only an aspect of it.
Few translation theoreticians have applied the insights of formal game theory to translation.A notable exception from this is Jiři Levỳ, who attempted to construct a formal model for the decision process in translation (Levỳ, 1967).Levỳ of course did not think that translation would allow to develop strategies that would work if not in all, than in almost all cases, but he suggested that translators solve their tasks according to the minimax principle.He argued that translation theory tends to be normative, to instruct translators on the optimal solution; actual translation work, however, is pragmatic: the translator resolves for one of the possible solutions which promises a maximum of effect with a minimum of effort.That is to say, he intuitively resolves to the MINIMAX STRATEGY (Levỳ, 1967).
We agree with Michael Cronin, who believes that the shortcomings of Levỳ's theory are partly those of a formal game theory itself (Cronin, 2001).The assumption that players, in our case translators, act rationally is constantly contradicted by various factors that determine translation choices: stress, speed, fatigue etc.Quantification of those factors, in our view, is a major difficulty for any formal game theoretic representation of translation process.Plus, the issue of imperfect information is not properly dealt with.Game theory deals with the issue of imperfect information by assigning probabilities to alternatives and establishing pay-off matrices.However, in translation such matrices quickly become extremely complex.It is only in the presence of a restricted vocabulary and syntax and in a sharply restricted text type that optimal strategies may be developed.Such strategies might allow a machine to produce a translation that would be comparable in quality with the translation produced by humans.
The notion of strategy, as used in games, needs to be somewhat refined in translation theory.If reading a text is a dialogical activity and texts are partly generated by reader's interpretive strategies, then where is the translator here?Are translators model readers (in U. Eco's terminology) or are they just a particular kind of readers, maybe even somehow privileged?
We propose to study translation not only in the light of game theory, but also of conceptual integration theory, proposed by Gilles Fauconnier and Mark Turner.The theory is based on the idea of mental spaces, defined as follows: partial structures that proliferate when we think and talk, allowing a fine-grained partitioning of our discourse and knowledge structures (Fauconnier, 1997).
Correspondences between mental spaces are of a complex nature and can be elaborated by means of information that does not originate from spaces in question, but can also stem from conceptual domains related to such mental spaces.Therefore, the internal structure of mental spaces is said to be motivated in both a conceptual and experiential manner (Fauconnier & Turner, 2001).
The associations between various mental spaces, called mappings, are activated when an individual conceptualizes the surrounding world.They are culturally and lexically rooted, hence they lie at the heart of semantic comprehension, language interpretation and mental constructions (Fauconnier & Sweetser, 1996).
Fauconnier and Turner postulate that conceptual integration theory ought to be treated as a universally acknowledged basic process which underlies human thought and language (Fauconnier & Turner, 2008).Despite the fact that blending mental spaces has become a prevailing theory in many areas of study, which certainly qualifies as an asset, the proponents of the theory have also suggested that it should be rendered as a simple cognitive procedure, applicable not only to complex scientific issues, but first and foremost to everyday processes of human understanding.Clearly, the human ability to produce texts in different languages is one such fundamental cognitive process as well.
The author of a text, being able to handle a specific set of mental spaces (for more details see for example G. Fauconnier & M. Turner (2001)) begins to construct a mixed space, which is in fact a semantic network.The number of mental spaces handled depends on the individual characteristics of the discourse personality.Mental spaces materialize in the mind when we think or say something.As a result of the creative process there appears the text (a product of author's discourse).Now, the goal of translator as player is to identify basic spaces and reconstruct them (the term reverse engineering reflects the essence of the process quite accurately).
Our further ideas are based on observations of the work of fellow translators, as well as the work of students in the online club The Craft of Language, which is conducted under the same name of the social network Facebook. 419 students of the first year Master of Moldova State University, major in English philology, were observed.Students were asked to translated small texts (around 10 lines) of different genres from English into Russian and Romanian.
The game of translation for any kind of text, following the idea expressed by E. Kunitsyna in the context of the study of translations Shakespeare as game, can be divided into two components: author-translator, translator-audience (Куницына, 2009).One must clearly distinguish between these two stages because there are cases when, after reading or hearing a phrase in any language it is not always possible to immediately convey its meaning, even by means of the native language.Those stages were clearly visible in the activities of students that we have observed in our sessions.They first needed to understand the message and only after that could they proceed to formulating it in Russian or Romanian, as they were asked.If there is experience of working with a particular type of text, the game author-translator may not be visible at all, as many procedures are habitual and automatized.
It is also necessary to distinguish between translation as equivalent substitution and translation-interpretation as a search for the appropriate equivalent using special knowledge (about the topic of messages, about the culture of native speakers, means to express a particular idea about the context of communication, etc.).
During the observation of Translation Braintwister, a weekly online translation club where master degree students from the State University of Moldova participate, 5we have clearly noticed the existence of the two stages.The students were unable to produce a good translation until they would clearly understand the meaning of the author's discourse.If understanding was achieved in all cases the students were able to correct the statement in the target language on their own, without any assistance.
The task of the translator in the game of strategy that she plays is rather tricky as the translation must fit into the family of similar texts, initially written in the target language.

Translator-homo ludens
The translator as well as the representative of any other profession, is subject to evaluation.His or her personal qualities as well as the results of the performed activity are evaluated.The capacity of being subject to evaluation allows us the assertion that the phenomenon of reality is basis for the development of the concept (Арутюнова, 1998; Слышкин, 2000).
While studying the concept of translator, we have identified an increased level of semiotic density that, in our opinion is in tight relation to the relevance, importance of this concept in the eyes of the cultural and linguistic community, to the axiological and theoretical value of the phenomenon reflected in its content.We would like to mention that the concept of translator cannot be reduced to a certain language or to a certain culture due to the particularity of the developed activity that is the process of translation.
The semiotic density of the concept of translator can be partially explained by the array of metaphors it comprises.Moreover, it also has an emphatic shade of playfulness.The element of playfulness cannot be discharged from speech and is related to the fact that for speakers of a certain language it is not just the thought that is important, but also the way it is expressed, and it's a fact that there are several metaphors within the concept of translator, which implies risk and therefore a game as well.
According to J. Huizinga, risk is one of the main attributes of the game (Huizinga, 1949).What risks are there in the process of translating?First of all there is the actual danger related to the activity of the translator for example during warfare.In the open letter signed by the international professional associations of translators AIIC, FIT, IAPTI, addressed to the Prime Minister of the United Kingdom David Cameron on 06.13.2013, it is stated that translators who assisted the British troops have become main targets for militants, and during April-June 2013 alone 25 representatives of this profession have lost their lives. 6A translator could risk his life working on a text that is prohibited out of political, religious or other reasons.The translators of the works of Salman Rushdie have become victims of terrorist attacks. 7The translations of Shakespeare's works have at times been censored, while Macbeth and Henry IV have even been banned in royal and demotic theaters.Therefore, risk is an ontological component of the concept of translator and one of the determinants of the intentional horizon of the translator.
The well-known opinion of W. von Humboldt depicting the translator as a navigator, serves as confirmation of the fact that the translator does take certain risks.Every translation is for him or her, an attempt to solve a matter that has no solution.Any translator will inevitably hit underwater cliffs if he's very loyal to the original because of the tastes and language of his or her own people and the same will happen if he or she tries to preserve the characteristic of the nation he or she represents, on account of the original.
In the works of I. Zubanova there are three metaphors based on the idea of risk.The author compares the interpreter to a juggler, an air traffic controller and a sapper.Just like the air traffic controller, the translator has to multitask, follow the development several cognitive processes (his or her own and those of the speaker), forecast the subsequent events and try to match the number of take offs with the number of landings, that is to ensure that the translation mirrors the speaker's ideas and emotions to the full.Just like the sapper, the translator isn't entitled to mistakes.In our opinion this aspect is arguable, but we won't go into details because this discussion would exceed the aim of our study.The translator-juggler sometimes risks not catching the fragile object that is the meaning and the signification (Зубанова, 2004).
Another interesting metaphor is the one A. Ghitovici, poet and translator from Chinese refers to.He compares the translation of the works of the Eastern classics to climbing an extremely high mountain, a task that can be performed only by an experienced, judicious and brave alpinist (Гитович, 1969).
The translator is the one who takes care of the borders of the text and goes beyond them, if needed.On this train of thought it would be interesting to recall a quote from the letter of P. Antokolski to B. Pasternak, about the translation of "Romeo and Juliette", done by the latter.Antokolski speaks about the fact that Pasternak chose to perform that very translation in order to cross the borders of dictionaries and offer new feelings and experiences to the foreign actors and audience (cited in Куницына, 2009).
Yet another interesting metaphor is the one representing the translator as a slave and a competitor.The famous Russian poet V. Jukovskii said that when it comes to fiction the translator is a slave and when it comes to poetry, the translator is a competitor, while A. Schlegel compared the translator and the author to the participants in a duel, that ends with a downfall for one of them (Топер, 2001).
We consider it more appropriate to compare the activity of the translator to that of an adroit perfumer, who can identify the components of a perfume and create new scents.
The fact that there are elements associated with risk within the structure of the concept of translator confirms the rectitude of our idea to present the process of translation in terms of game theory, and the translator as homo ludens.
The risk is one of the factors that determine the intentional horizon of the translator.In terms of cognitive theory of translation we can assert that the translator runs the risk of not managing to balance two main intentions: correspond to the structure and correspond to the goal (or experiences).
From the phenomenological point of view we have sufficient arguments to assert that the translator is a player, a homo ludens.Among all states of intention, there is one that deserves particular consideration.And that one is doubt.It is namely this feeling that underlies the cognitive dissonance of the translator, which provides for the process of overcoming the disparities among languages and cultures.The degree of cognitive dissonance determines the level of proficiency in translation (Воскобойник, 2004).It is namely the cognitive dissonance that measures the professional excellence of the translator.A constructive perfectionism is an indispensable feature of the translator who cannot and must not be absolutely sure of the rectitude of his translation.He/she is sure there is an alternative way to translate the text he/she is working on, that could even be better than the one suggested, but risks anyway and goes ahead with his/her version, even if he/she will improve on it afterward.
It is namely the risk and the cognitive dissonance that form the elements that allow us to present the translation process as a game.

Qualitative Study of Translation Process Perception Inside and Outside Professional Translators' Community
We are interested in the way that professional translators, as well as people who do not practice this activity, perceive the process, the qualities necessary for it.For the purpose of collecting empirical data we used the on-line survey method.We have chosen this method because it:  allows diversification of the sample of respondents with little effort;  makes it easier to ensure the objectiveness of the obtained data;  facilitates the task of ensuring the protection of personal data of respondents (provided by the laws of the Republic of Moldova).
In order to be able to place the survey on the Internet we created a user account with the online survey tool www.surveymonkey.com.The survey was generated with the help of the tools embeded in the respective service and had a permanent link, 8 that was used as a data collector.This link was placed on the wall of the Facebook social network group The Craft of Language, managed by the author. 9The members of the group are professionals from the field of translation from the Republic of Moldova, State University of Moldova professors, master's students of the department of foreign languages and literature from the State University.The link was sent to our fellow translators and researchers via email and placed on a forum for professional translators on one of the most important professional sites proz.com.We also used the function Submit a Query available on the linguistlist.orgportal.By filling in the form and making the link to the form available to others, the researcher can obtain answers from the necessary respondents.It is possible to submit the query on the following conditions:  Linguistic relevance: the answer to the question asked must concern linguistics, and not literature or computational aspects of language study.


Global relevance: the question asked must regard linguists from any part of the world, without confinement to a certain region.


Academic relevance: the message must be of interest to the general linguistic community, and the one who initiates the survey or the question must be well-grounded in the addressed matter.Elementary, redundant questions are not allowed as well as those the answer to which can easily be found using search engines.
Observance of this conditions is monitored by a moderator.Only after the message is approved can it be sent to the community of linguists subscribed to the specialized mailing list.
Thus we managed to ensure a high level of diversity among respondents as well as save a lot of time.The development of the online survey via surveymonkey.comservice was in line with the legal provisions of the Republic of Moldova regarding the protection of personal data of participants to scientific studies.The questionnaire did not request information that would allow identification of the respondent.All respondents decided on their own regarding the amount and type of data they wanted to provide in answer to the questions in the questionnaire.We did not consider it necessary to request that every respondent sign a declaration of agreement to participate in our survey.Accessing the link is considered tacitly expressed agreement to participate.
The surveymonkey.comservice allows the researcher to edit the respondents' answers.The rationale behind it is unknown to us.Any modifications are immediately reflected by the software used for displaying the answers, with the indication of date and time of the modification.
We have not made any modifications to the provided answers.It was not necessary.
An example of the way in which the respondents' answers were displayed is provided in the next page.
The use of surveymonkey.comservice allowed us to ensure the sufficient level of confirmation of provided answers authenticity.For each answer, the software provides information regarding the questionnaire operation inception and conclusion time, the used collector (the way in which the person reached the page of the survey) and the used IP address.We can easily obtain information about the country from which the respondent accessed the questionnaire with the help of the free service http://whoissearch.com(see the next page).According to the law, this information does not constitute personal data, as it does not allow the identification of the respondent, therefore it can be used for the purpose of the undertaken research.
There were a total of 77 respondents from different countries, including the Republic of Moldova, Germany, Uruguay, Italy, Croatia, Slovakia, Chile, Russia, Great Britain, France, Argentina, Japan, Indonesia, Canada, Romania, Brazil, Czech Republic, Portugal, Cyprus and Israel.
At this point we consider it necessary to make a comment and explain some aspects referring to the number of respondents for a qualitative study.The sample for a qualitative study is usually much smaller than the one for a quantitative study.When working with a sample from a qualitative point of view, it is not always the case that extra data generates extra information.This roots its explanation in the fact that the occurrence of a certain data fragment or immediate code leads to its inclusion in the analysis framework.Frequency is rarely important in qualitative research.Information or data that occurs just once or repeatedly might be equally important to the understanding of processes that stand behind it, for qualitative research is based on the identification of the exact meaning, not on the statement of some generalized hypotheses (Crouch & McKenzie, 2006).Moreover, studying a subject or object from the qualitative point of view demands a lot of work and the analysis of a larger sample could be timeconsuming and not practical at all.
The participants may be different in any field of studies.
The sample for a qualitative study must be large enough to ensure that all perceptions or their largest part that could be potentially important is covered, but at the same time, if the sample is too large, the data begins to repeat itself and eventually becomes redundant.If the researcher remains loyal to the principles of qualitative study, as a rule, the size of the sample is determined by the so-called saturation (Glaser & Strauss, 2009), this is to say that collecting new data will not provide extra information on the studied problem.
Our study does not focus on ethnography and ethnology, or grounded theory, and it could only partially refer to phenomenological research, in the sense that it is an attempt to describe a range of phenomena within the translation process, so we will go by the recommendation that fifteen is the smallest admissible number of respondents.Thus, our sample is large enough in terms of quantitative parameters.Some researchers will even say it's too large.The fact that our questions have been answered by translators of diverse origins who operate with different languages, allows us to state that the sample of respondents was sufficiently diversified.
The answers provided by translators have been processed using the special program ATLAS.ti, for managing data from qualitative surveys.A screenshot showing an image of the used program is provided below.ATLAS.ti is an excellent tool for qualitative analysis of large fragments of textual data, images, audio and video recordings.The researcher has at his or her disposal various means of accomplishing any task connected with the systematic analysis of unstructured data, that is to say empirical data that does not comply with statistical and perfunctory means analysis.The toolbar provided by ATLAS.tienables the management, extraction, comparison and reassemblage of huge fragments of data in a creative, flexible and systematic way.
During the analysis descriptive, procedural and in vivo codes have been generated.
It turned out that 26 people (33,8%) out of a total of 77 respondents indicated they were not translators.But there is a remark to be made.The respondents probably referred to the traditional interpretation of the term translator-the person who performs the translation of written discourse.Among participants there were persons who are not in the traditional sense of the term, but who practise or practised translation of spoken discourse.All the respondents know at least two foreign languages.
The asynchronous, in-depth interview contained the following questions: 1. Are you a translator or interpreter?The questions were asked in three languages: English, Romanian, and Russian.Answers were allowed in English, Romanian, Russian, French and German (languages which the author knows).The respondents answered in English, Romanian and Russian.The vast majority of the respondents chose to answer in English.
Mention must be made, that the article presents a part of a broader study and in this paper we shall provide detailed analysis of responses to two questions only, which allow us to come up with a better picture of translator as homoludens and explain the need for further investigation of the issue from the sociolinguistic point of view too.
When asked to choose the types of knowledge most essential to a translator, the respondents were given the possibility to choose multiple options.They were also given the possibility to comment on their choice.48 people decided to comment.The distribution of the respondents' choices is shown schematically in Figure 1.
It is interesting to observe that the option "knowledge of mother tongue" received the highest number of "votes" from respondents (54).One of the respondents even stated that translator's good knowledge of mother tongue is sometimes underestimated: Mother tongue knowledge is vastly underrated, and yet is the most crucial limiting factor on the quality of the end product.
Knowledge of foreign language received 50 votes.The knowledge of culture received 46 votes from the linguists who participated in our survey, ranking in third.32 people considered that it is very important for a translator to be familiar with the field of every particular translation, that is to possess technical knowledge.It seems that practitioners' opinion doesn't match the one of the theoreticians of translation who increasingly claim that it is extremely important for a translator to be bicultural, to know the subtleties of the cultures that come into contact during verbal communication, and even spend a good part of his or her life poising on the border between the cultures of the speakers with whom the translator works.
It is interesting to observe how the profession of translator is perceived among those who speak at least one foreign language, but do not consider themselves translators.Nearly all respondents (19 out of 26 people, who do not consider themselves translators) think that the most essential knowledge is knowledge of foreign language and knowledge of the culture of speakers of that particular language.It is interesting that beside the somewhat standard answer to the option something else, that the translator is a complex person and enjoys various types of knowledge, one of the respondents underlines the importance of computer literacy: At least basic computer skills in today's market.
It is extremely interesting to observe the way the mental representation of the translator is developed for those who do not belong to this profession.The particularly outlined image corresponds to the one shaped after studying the translation discourse from multiple textbooks and theoretical studies regarding translation, though written by those who practise this profession occasionally or do not practise it at all.This occupation is perceived by others as something extremely difficult, that requires much effort.Thus, one of the respondents stated: Steady nerves, being in the middle of a conversation can be tense and it's difficult to keep up and be accurate in both the words and the meaning that are being conveyed.
The translator is perceived as a special person even from the biological point of view.One of the respondents stated: [Translators have] brains that can do things that mine can't.
The opinion of another respondent, that is worth mentioning is as follows: Translators need profound knowledge of stylistic and other connotations that expressions carry, which requires high proficiency in both the source and target language as well as familiarity with all the cultures involved.Interpreters need a brain that is adapted to simultaneous processing and production of different language.Faithfulness with respect to anything other than truth conditions seems to be a secondary concern here.
The same distinction which persists in the speciality literature, concerning the necessary skills for the translation of written and spoken discourse can be noticed here, as well as a totally wrong opinion concerning the loyalty to the original, in what regards the translation of spoken discourse/oral speech.The experience of the author of this work, as a conference interpreter, shows that text processing does not occur simultaneously and that it is actually impossible.The translation of oral speech does not focus exclusively on the principle of loyalty to the original.An elaboration of this subject would go beyond the scope of this work and may become subject of a separate study.

Conclusions
Translation studies and teaching of translation should not and must not be limited to mere analysis of parallel texts, as there is much more lurking behind the result of the process.The study of the infinite game of incomplete information, which is translation in the broad sense of the term, is only at its beginnings.
The process of translation is perceived very differently inside and outside translators' community, as the responses analysed in present paper show.What needs to be mentioned at this point is the fact that very often the point of view of the outsider to the profession is found in books and papers that study translation and, what is even worse, it is reflected in the curriculum structure of courses designed to teach translators.Gaining a better understanding of the process of translation, not of its product, will eventually allow not only to educate better translation professionals, but also improve the algorithms of machine translation systems.
Semiotics, game theory and decision-making theory taken together seem to offer the best instruments to analyse the issues of translator as homo ludens.The fact that there are elements associated with risk within the structure of the concept of translator confirms the rectitude of our idea to present the process of translation in terms of game theory, and the translator as homo ludens.Risk is also the key element of translator's episteme and the determining factor of translator's intentional horizon.

Figure 1 .
Figure 1.Essential types of knowledge from the point of view of those who do not consider themselves translators

Figure 2 .
Figure 2. "Essential knowledge" from the point of view of those who do not consider themselves translators