Using at least one example of cross-disciplinary research in Cognitive Science, illustrate and explain the idea of interdisciplinarity

Introduction

 An interdisciplinary subject is a subject that uses research from a variety of different academic disciplines to create its own theories. Cognitive science is known to be one of such disciplines. Cognitive science is to uncover the fundamental principles that all natural and artificially intelligent life abide by, from high-level reasoning to motor behaviour. This science synthesises characteristics from many different fields such as Artificial intelligence, where the studies have been carried out to understand how the human brain learns and whether that knowledge can be used when creating Machine, so it can be an artificially intelligent system. Psychology is another discipline where cognitive science can draw inspiration from. Using studies into the behaviour of the mind, this includes the studies into the unconscious and conscious mind. These studies help cognitive scientists to achieve their goal as insight is provided into the Physiological processes that the brain carries out in respect to memory, learning and language. The field of cognitive science is boosted with the help of the neuroscience discipline. This takes a biological approach to how the brain carries out its processes, such as how the nervous system link up all the different stimuli that are received and how this information is passed to the relevant parts of the brain for storage, likewise with psychology , neuroscience also look into how the brain stores different stimuli and how it is processed to produce a result. i.e. Speaking a language. On top of this, the study of philosophy is beneficial to the science as it asks very important questions like “what is consciousness? “, “if a system can become conscious and if so how does it decide what stimuli are worth remembering”.

Main body

Philosophy has played a key role in the study of Cognitive science. There has been a merger between the two subjects creating a new strand called cognitive Philosophy in which questions such as ‘What is consciousness’ can be asked. This along with other related questions such as ‘what is free will’ have yet to be answered although there are studies that have helped philosophers get closer to a definitive definition.  initially the idea of “Cartesian Materialism” (Dennett, 1991) This is the idea where various parts of the brain unconsciously begin to process different stimuli. The processed information is then sent to a focal point where everything comes together for the homunculus; this focal point is referred to as the Cartesian theatre or inner self.  (Dennett D. , 1996). Although there is no evidence that such a place exists, many philosophers concluded that something like that must occur for consciousness to be able to store memories. However, this then presented a further question” is the inner man conscious? and how does it know it is conscious?” this question posed a lot of issues there was no comprehensive answer that doesn’t involve going around in an infinite loop about what can make the decision regarding what is focused on and what isn’t. This led to another thesis created by Daniel Dennett’s called the multiple drafts metaphor. According to Dennett, the conscious state spread out across time and space, it is spread out in the brain across multiple content fixations. Each fixation is a draft these drafts compete for dominance in the cognitive system. This is known as “Fame in the brain “. This metaphor is supported by the colour phi test. This is where a subject is shown a green flash and then a red one. the results showed that the subject had experienced a motion between the two flashes where the green circle changed to red at the halfway point. Some explanations to this phenomenon included “the subject unconsciously predicted the red flash thereby creating a false memory of the flashes moving.” However, this was disproved by Dennett as there is no basis for choosing an explanation as it is down to how the subject perceived the test. This shows that there are many different drafts that all compete in the cognitive system. This metaphor came under criticism too as it goes against verificationism. Without a philosophical understanding cognitive science would not be able to move forward as there would be too many accepted theories that would contradict each other.

For some time, psychology has played a major impact on the development of artificial intelligence. The goal for artificial intelligence is to create a system that can learn and adapt to a situation without the input of a human. Currently, there is no such system that can really be called AI yet. This has led the development of AI to be kept into two different sections. Weak Ai is most of the AI that we have today, this type of AI is purpose-built where it can only handle the specific task in hand. Strong AI is thought to be the future of AI, this type is supposed to carry out cognitive tasks better than humans. Research into whether machines can be intelligent has a long history, Alan Turing devised a test called the Turing test (A.Turing, 1950 ). This composed of a human as a question master and two respondents; a computer and a human. The aim of the test is to fool the question master into thinking that the computer is a human. This test too has also come under heavy criticism as if the questions must be limited to true-false to allow the computer to exhibit a human-like response. John Searle a philosopher has an opposing view on the possibility of strong AI. Searle proposed the Chinese room experiment (J.Searle, 1980). In this test nor the human or machine truly understood the task in in hand but were able to complete the task proficiently. This contradicts the Turing test because in both cases the machine still showed human-like responses.

 There are problems with the way Strong AI learns. this is due to the humans trying to make an artificially intelligent system like the human brain. humans have many cognitive biases and this bias will influence the AI though input stimuli (Rosso, 2018). this means that the type of data that fed to the system will affect AI. An example of negative data that had been given to an AI is Microsoft Tay(bot). this system mimicked the language patterns of a nineteen-year-old girl and was given the ability to learn from interacting with other Twitter users. Shortly after its release many of the users started to feed the AI’s knowledge bank with politically incorrect and offensive statements that cause the chatbot to be taken down, as the bot began tweeting things from its knowledge bank. Recent examples of this show even though AI can learn it still lacks the reasoning and understanding of what is being tweeted.

Likewise, with Psychology, Neuroscience also has strong links to cognitive science creating a separate strand called Cognitive neuroscience. This discipline follows a more biological approach to researching the brain to find answers concerning the physical mechanisms to do with behaviour. Cognitive science uses both invasive techniques to measure brain activity in the form of knockout mice where certain parts of the mice’s brains are switched off and non-invasive techniques such as fMRI where the brain is scanned and areas with significant activity are highlighted

Cognitive linguistics and cognitive Neuroscience have been used in conjunction with each other. There have been numerous studies into the way languages are perceived and learnt. From early research into patients with brain lesions doctors were able to discover to key brain regions; the Broca’s region and the Wernicke’s region. Since then researchers have found ways to find other areas that are linked to language. The Wada test (Wada, 1949) is an example of this. In this test patients are injected with a barbiturate through the carotid artery, this puts half of the brain to sleep. This allows doctors to run different cognitive tests to see what the subject can or cannot do. This test shows that most subjects process language in the dominant hemisphere. Some more recent imaging studies show that language processing is split into a duel stream process. This information is sent along the dorsal pathway which helps to control movements; for example, mouth movements, the information is then sent along the ventral stream; this process what is being said and in what language. Other research into language learning has busted popular misconceptions that were thought to be true. (Nouri, 2015) A popular myth was that if children were exposed to a foreign language at an early age then it would impair their knowledge of their first language. However, the reality is that Learning a foreign language helps to foster other languages and allows the individual to be more competent towards other languages. Although this claim is not supported by the various scientific studies as many of the studies that were conducted in the early 20^th century had faulty methodologies. Where the subjects that were chosen were from different cultural and social conditions. Many of the foreign subjects did not have a sturdy base in their own native language making learning another language more difficult, compared to subjects from privileged environments with a sturdy base in their home language.

Computer science also has an interdisciplinary relationship with cognitive science too. Artificial intelligence has taken many of the findings from cognitive science to get closer to creating an artificially intelligent system. The idea of Artificial intelligence is very broad and has many subsets; for example, Machine learning. Machine learning is where the system is exposed to big data sets. This data is then run through numerous algorithms until it achieves a similar result, the algorithm is then tweaked so that the result is completely correct. A collection of these different algorithms come together to form a neural network.  These networks are like how the human brain works, therefore, its research into these networks may unlock the secrets of the human brain and vice versa. Deep learning is a subset of machine learning and is regarded to be the future in AI learning. Whilst machine learning systems use one layer of “functions” to reach the required output, deep learning systems use multiple layers to create an artificial network. These two approaches to AI need to be used in tandem with each other. For deep learning to be efficient large data sets are required for the system to understand the task. An example of where deep learning AI has been used in society is in the process of adding sound to silent movies. This AI is trained using over a thousand examples of videos. The deep learning model will associate the sound frame by frame with different sounds from its database. It will then select the best sound matching the scene. (Patidar, 2018)

Conclusion

As we can see the subject of cognitive science has benefited greatly by using research from the other academic disciplines. This use has caused the subject to take great strides in recent years, the subject has found common ground however some challenging theories between the disciplines remain.  For example, the idea of consciousness. The philosophical questions Surrounding this topic are yet to be answered. This shows that the subject is still a long way from this interdisciplinary cohesion “we are still a long way for understanding how cognitive processes actually work.” (Stuart Russell, 2018)Until full understanding is achieved there can be no single theory into how a thing can become conscious. The main problem with AI is that it is still way too difficult for Human intelligence to be recreated. Some of the simplest tasks such as communication is too complex to code as the machine will have to understand the natural language, context and tone of what is being said. In addition to this it will have to think of an answer from its knowledge bank and then produce an answer to in the same natural language that the recipient can understand. This issue will not be around for the next 50 years as there are companies that are getting closer to achieving human-like intelligence.

Works Cited

• A.Turing, 1950 . Computing Machinery and Intelligence, Manchester: s.n.
• Data Flair, 2018. Deep learning Vs Machine Learning. [Online]
  Available at: https://data-flair.training/blogs/deep-learning-vs-machine-learning/
• Dennett, D., 1991. Consciousness Explained, s.l.: s.n.
• Dennett, D., 1996. Consciousness: More like Fame than Television, s.l.: s.n.
• Emerj, 2018. What is Machine Learning?. [Online]
  Available at: https://emerj.com/ai-glossary-terms/what-is-machine-learning/
• Engadget, n.d. Machines can generate sounds effects that can fool humans. [Online]
  Available at: https://www.engadget.com/2016/06/13/machines-can-generate-sound-effects-that-fool-humans/
• Gennaro, R. J., n.d. [Online]
  Available at: https://www.iep.utm.edu/consciou/#SH4c
• J.Searle, 1980. Minds, Brains, and Programs, s.l.: Behavioral and Brain Sciences.
• Nouri, A., 2015. Cognitive Neuroscience of Foreign Language Education: Myths and Realities. Iranian Journal of Research in English Language Teaching, s.l.: s.n.
• Patidar, S., 2018. Machine Learning vs. Deep Learning. [Online]
  Available at: https://dzone.com/articles/comparison-between-deep-learning-vs-machine-learni
• Rosso, C., 2018. The Human Bias in the AI Machine. [Online]
  Available at: https://www.psychologytoday.com/gb/blog/the-future-brain/201802/the-human-bias-in-the-ai-machine
  [Accessed 14 january 2019].
• Stuart Russell, P., 2018. Artificial Intelligence: A Modern Approach. 3rd ed. s.l.:s.n.
• Wada, J. A., 1949. A new method for the determination of the side of cerebral speech dominance. A preliminary report of the intra-carotid injection of sodium amytal in man, Tokyo: s.n.