CC BY
8https://doi.org/10.48417/technolang.2022.01.08 Research article
The Quasi-Other as a Sobject
Larissa Ullmann (E) Darmstadt Technical University, Karolinenpl. 5, Darmstadt, 64289, Germany, [email protected]
Abstract
This commentary concerns the concrete use of linguistic terms to describe the technical other, the robot, and its relationship to humans. There are many characteristics that a robot can have that are very similar to humans and interpersonal relations, but they are not human, they are quasi-human. This phenomenon is, amongst others, constructed and interpreted linguistically, but on the other hand, there is no linguistic term that could describe it unambiguously, so it can only be studied in direct human comparison, in a quasihuman way. In this comment, it is demonstrated why the use of the quasi is problematic and suggests that the phenomenon can instead be analyzed in a techno-philosophical-phenomenological context within the framework of the Sobject-approach The term sobject describes a kind of technical objects to which humans can have deeper relations than to conventional objects. Therefore, it provides space to study the phenomenon on a phenomenological level, without the need for a permanent direct human comparison. -This is one of six commentaries on a 2011-paper by Mark Coeckelbergh: "You, robot: on the linguistic construction of artificial others." Coeckelbergh's response also appears in this issue of Technology and Language.
Keywords: Human-robot relation; Language; Phenomenology; Technical others; Objects and sobjects; Artificial Intelligence; Hermeneutics
Citation: Ullmann, L. (2022). The quasi-other as a Sobject. Technology and Language, 3(1), 76-81. https://doi.org/10.48417/technolang.2022.01.08
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
УДК 1: 62-529
https://doi.org/10.48417/technolang.2022.01.08 Научная статья
Квазидругой как собъект
Лариса Ульманн (И) Дармштадский технический университет, Каролиненплац 5, Дармштадт, 64289, Германия
Аннотация
Данный комментарий к тексту Марка Кекельберга "Ты, робот: о лингвистическом конструировании искусственных других" касается конкретного использования лингвистических терминов для описания технического другого, робота, и его отношения к людям. У робота может быть много характеристик, которые очень похожи на человеческие и межличностные отношения, но они не человеческие, они квазичеловеческие. Это явление, среди прочего, конструируется и интерпретируется лингвистически, но, с другой стороны, нет лингвистического термина, который мог бы описать его однозначно, поэтому его можно изучать только в прямом человеческом сравнении, квазичеловеческим образом. В этом комментарии показано, почему использование "квази" проблематично, и предполагается, что вместо этого феномен может быть проанализирован в техно-философско-феноменологическом контексте в рамках объектного подхода. Термин "собъект" описывает своего рода технические объекты, с которыми люди могут иметь более глубокие отношения, чем с обычными объектами. Таким образом, он предоставляет пространство для изучения явления на феноменологическом уровне без необходимости постоянного прямого человеческого сравнения. - Это один из шести комментариев к статье 2011 года Марка Кекельберга: "Ты, робот: о лингвистическом конструировании искусственных других". Ответ Кекельберга также опубликован в этом выпуске журнала "Технологии и язык".
Ключевые слова: Отношения человека и робота; Язык; Феноменология; Технические другие; Объекты и собъекты; Искусственный интеллект; Герменевтика
Для цитирования: Ullmann L. The quasi-other as a Sobject // Technology and Language. 2022. № 3(1). P. 76-81. https://doi.org/10.48417/technolang.2022.01.08
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
INTRODUCTION: A MISSING TERM FOR THE QUASI OTHER-
In his article about the linguistic construction of artificial others, Mark Coeckelbergh (2011) talks about two "linguistic-phenomenological 'glasses' or repertoires" (p. 63) to describe the others or their relations to humans. On the one hand, there is an ontological view, which strictly separates subject and object and labels every technique, every robot as an object, a thing, while humans are clearly subjects. But on the other hand, we have a social ontology that allows a kind of ontological hybridity of human-robot relations (Coeckelbergh, 2011, p. 63). This second approach and the perception of a hybridity seems to be fruitful and important because it provides a space for discussion about a new kind of phenomenon. To describe this phenomenon of an other and the relation to humans, Coeckelbergh uses the following linguistic terms: the quasi-other, the artificial other, the artificial companion, a quasi-objective reality, a quasi-subject, and a quasi-social relation. These terms describe a phenomenon that allows a relation between humans and robots that goes beyond the relationship with conventional objects. We see that it is necessary to describe this phenomenon in linguistic terms, but it is unclear which words are the right ones or which are more suitable than others. Based on Coeckelbergh's linguistic-hermeneutic analysis, I would like to focus on this novel phenomenon, which has become increasingly important in recent years.
A social-phenomenological linguistic-hermeneutic approach provides space to philosophically investigate the phenomenon of quasi-others and quasi-social relations. But what does the quasi mean in these phrases? It seems like the quasi represents something real (in the human sense), but it's just not really real because it's technical and not human. There are human-like interactions with the quasi-other, so we can't just speak of the object or the other. This shows that we need language to describe the phenomenon, but also that linguistic terms are not sufficient. We always have to speak of quasi or human-like and can describe the technical possibilities exclusively in comparison to humans. However, the abilities and characteristics of quasi-others are often not comparable to those of humans. And here the problem also arises because the linguistic terms are not sufficient. Let us take the example of the terms think or intelligent. We describe technical intelligence as artificially intelligent, which often leads to the claims that artificial intelligence is better compared to human intelligence. This involves, for example, comparing the ability to calculate and concluding that machines could calculate better. In the context of artificially intelligent machines, it is also said that the machines could think faster or better when processing data. But machines and humans are not better or worse, they simply have a different way of thinking respectively processing. There are clear differences between artificial and human intelligence, mainly in the way they arise. But in the way of appearance and results, the two types can be very similar or even the same. This can also be seen in Searle's Chinese Room example or the Turing Test, which Coeckelbergh also talks about (2011, p. 64). But the crucial factor at this point is not only
PHENOMENON
THE PROBLEM WITH THE QUASI
the way we speak about and to robots and robot-human relations, but also that linguistic terms are missing to describe this quasi-phenomenon. So, on the one hand, we have a phenomenon that arises from the use of language and the way we talk about and to robots. On the other hand, there are these human-like quasi-relationships, and there is a lack of appropriate linguistic terms to describe and recognize these. Therefore, it is important to consider what the phenomenon is concretely. And that is why it can be helpful to reference this with a new, phenomenological, techno-philosophical term. The use of language co-constructs and co-interprets this quasi-phenomenon, but then this social-phenomenological phenomenon is there and we have to deal with it. I propose the term sobject to describe the relationships between human subjects and this specific kind of technical other from a phenomenological perspective.
A PHENOMENOLOGICAL SOLUTION: THE SOBJECT INSTEAD OF QUASI
The sobject can have relationships with humans that are similar to those with other humans and are especially different from conventional subject-object relations. It is based on Simone de Beauvoir's phenomenological approach, in which she describes subjects as setting (setzend) themselves and opposing (entgegen-setzen) other objects (1949/2015). In contrast, objects cannot set themselves, but are set by the subject. In this way, objects are passive and subjects are active. This results in interrelations between subjects and objects.1 When Coeckelbergh speaks of the robot as a thing, this easily fits into these interrelations and the robot would be nothing more than an instrument for the human. It (or sometimes we would say "he" or "she") is an object and the human is the subject. But the robot-as-quasi-other does not fit so easily into this classification and this is why the quasi-phrase is used. Moreover, Coeckelbergh talks about social and quasi-social relations. So, there is a kind of technical object that can enter into a kind of social relationship with humans that goes beyond the relations to conventional objects. Using the term sobject, this type of relationship and quasi-other can be described. Coeckelbergh's (2011) text focuses on robots that can live with humans (sharing a form of life), perhaps do tasks for them, and interact with them in some way (2011). These (social) robots can actively oppose the human subject and thus enter into the described interrelations with them. However, since we are dealing with technical artefacts and not humans, the main differences between the subject and sobject are the way in which the human-like characteristics are created. As with intelligence or think, it can be stated here that the result or appearance of humans and robots can be the same or similar, but is generated differently in each case. The appearance is similar, and yet not the same. We can only perceive the form of life that we know and that we can describe linguistically. The proposal to use the word sobject does not serve to replace the quasi, but allows a discussion and investigation of the phenomenon on a phenomenological level. It creates
1 This approach also has many references to the phenomenological-existentialist tradition and shows similarities to the phenomenology of Merleau-Ponty, Levinas and Sartre. Even though objects can exist without subjects, the focus here is on subject-object relations, which are different from subject-subject relations (Ullmann, in press).
awareness of what already exists by mentioning it by name. We can recognize and acknowledge that, in addition to subjects and objects, there are also sobjects with certain qualities that are not to be understood exclusively in comparison with human characteristics. And that is the reason why there must be a separate and clear term for the phenomenon, allowing it to be differentiated from the human and from the objective things. Especially the comparison to humans is not sufficient, because robots and machines are not human and have other possibilities and characteristics. And to give space for the description of these special characteristics, we need a vocabulary that is adapted to them. In this way, a variety of problems can be solved or, at least, their focus can be shifted. For example, we would no longer have the problem of robots being smarter than humans, because they are not comparable and just different in that aspect. Also the ethical doubts regarding the sociality of robots would shift, since they could not be social or even ethical in the human sense, and there should not even be that claim. I would argue that there can be a social relationship between humans and machines, but the machines interact socially on a technical level that is different from the human level. So, they can't make ethical decisions in the human sense or make decisions in general. They have their own different way of doing such things. And we should always be aware of that. Coeckelbergh (2011) asks a question that goes in a similar direction: "We may try to imagine what it would mean if robots had their own way of doing things, if they developed their own form of (artificial) life" (p. 65). And with regard to ethical questions he assumes that in regard to them the phenomenon, the quasi-other, the sobject, already exists and is recognized, because otherwise these questions would not arise at all. Hence, the sobject can put the focus on the novel technical phenomenon and provides that this kind of own way of doing things and the own form of (artificial) life can be researched. This research would be phenomenological, but makes use of linguistic terms, which goes hand in hand with Coeckelbergh's general understanding of the relation between language and technology. Moreover, as he suggests, other questions arise from the phenomenological-linguistic investigation that should be discussed. Should we assign a gender to robots or (artificially intelligent) technologies in general? We already say "he" or "she" to them or even refer to ourselves and (our) sobjects as "we." With the help of the sobject-approach and a form of existence which is different from human and conventional objects, we can imagine introducing a new pronoun as well.
LANGUAGE AS A CONSTRUCT OF GENDER
To be sure, it is not only through the use of personal pronouns that we construct a reality in which robots have a gender. That this is more complex can be seen, for example, in the German language. Here, the term "der Roboter" is used to refer to robots in general, so that it is linguistically natural to speak of "ihm" (him) or "er" (his). But gender is constructed by much more, which manifests also in language. When robots are given a human-like appearance that is male or female, and are also given a male or female name, we assign a clear gender to this technical artefact. For example, the digital communication system Alexa is referred as "she" and her answers are "hers." Her femininity is constructed through the female name and voice, which is then reflected in language use. Thus, these
and similar problems possibly arise because the sobject category is not accepted. If it were accepted, perhaps there would be the possibility of not declaring techniques as female or male, but having their own term. As long as technology can be perceived, understood and interpreted only in direct comparison to humans, artificially intelligent social technologies will only be seen as human-like, but not human, instead quasi-human.
CONCLUSION
It becomes clear that there is a phenomenon that can be well described and shown on a phenomenological level, and which Coeckelbergh already points to through a linguistic-hermeneutic construction. However, the phenomenon is not only manifested by language and draws attention to the fact that we lack linguistic terms to deal with it. To overcome the limitations of our constructions and interpretations of reality, the Sobject-approach allows us to focus on relationships with robots and technological others in general, and also to describe them detached from immediate human comparison.
REFERENCES
Coeckelbergh, M. (2011). You, Robot: on the Linguistic Construction of Artificial
Others. AI & Society, 26, 61-69. https://doi.org/10.1007/s00146-010-0289-z De Beauvoir, S. (2015) The Second Sex (C. Borde, & S. Malovany-Chevallier, trans.).
Vintage Publishing. (Originally published 1949) Ullmann, L. (in press) Das Sobjekt: Mögliche Beziehungen zwischen Mensch und Maschine aus einem phänomenologischen Blickwinkel [The Sobject: Possible Relations between Humans and Machines from a Phenomenological Point of View]. Jahrbuch Technikphilosophie 2022, vol. 8. Nomos.
СВЕДЕНИЯ ОБ АВТОРЕ / THE AUTHOR
Лариса Ульманн, [email protected] Larissa Ullmann, [email protected]
Статья поступила 27 января 2022
одобрена после рецензирования 15 февраля 2022
принята к публикации 26 февраля 2022
Received: 27 January 2022 Revised: 15 February 2022 Accepted: 26 February 2022
