To what degree is the AI based Computer Linguistics comparable to Human Linguistics?
The cognitivist paradigm, dominant throughout the second half of the 20th century, drew a close parallel between computers and human minds, seeing cognition as the manipulation of abstract representations of external reality, by means of codified rules resembling software algorithms.
During the last quarter of a century, many researchers have moved away from this paradigm, adopting a transdisciplinary perspective drawing on cognitive science, phenomenology, and biology, seeing cognition as en emergent phenomemon that results from the interaction of an organism with its environment. In the case of humans, a sociocultural dimension is added to the physical environment, as we collectively weave complex patterns of meaning and intersubjectively project them onto ourselves and our surroundings. In this dynamic system, the role of language in human cognition and communication takes on a dimension that is far beyond what artificial intelligence can mimic. Beneath linguistic signs are complex patterns of multisensorial experiences resulting from the 'structural coupling' of a living body with its environment.
The emerging paradigm of embodied cognition makes it necessary to rethink the most fundamental aspects of conscious and unconscious thought, language, visual communication, aesthetic experience, etc.
I have been building a thematic bibliography about embodied cognition and aesthetic experience over the last three years, for my students in a seminar on this subject. Section 2 of this bibliography, "Cognitive studies, neuroscience, phenomenology, pragmatism, and semiotics," has the theoretical bases of the embodied paradigm (see particularly Chemero, Lakoff/Johnson, Shapiro, Varela/Thompson/Rosch, and Ward/Stapleton). Section 8 is dedicated to "Language, reading, writing, and literature" and has several publications that may be of interest to you (you might look at Cuccio, Cuccio/et al., Gallese/Cuccio, Gibbs/et al., Kravchenko, and Scarinzi). Here's the link:
https://www.researchgate.net/publication/317351468
I am looking forward to reading future answers to this question.
I agree with David and namely with the statement "... the role of language in human cognition takes on a dimension that is far beyond what artificial intelligence can mimic. "
My impression is that the computational paradigm in research on Natural Languages owes VERY MUCH to logics (although NOT to the "formal" linguistics) but NOT ENOUGH to the "functional" component of structural linguistics.
Let me add one curiosity however: a Japanese specialist of machine translation declared once that the failure in making a better automatic translator was actually due to the insufficiency of the transformational linguistics.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence