Is scientific knowledge the most reliable knowledge we have? (Essay vs Layman’s)

The following blog contains two versions of the same topic. Each paragraph will start with a referenced explanation and then a more digestible and public friendly explanation of what was just said in bold. If you are not familiar with how scientific literature works, scientific writing must build upon other referenced works in order to say anything of value. This usually makes the explanation squirrelly and horrible to read (some topics and writers are worse than others). Then again, summarising in a user friendly fashion loses accuracy.

The philosophy of science has been much debated over time, so to distinguish the reliability of the scientific method it requires a concrete definition. To demarcate science from pseudoscience, Karl Popper’s theory of falsification requires a theory to be able to be subjected to testing and modification, to raise the possibility of falsifying the hypothesis (Harlow, Cummings, & Aberasturi, 2007). While philosophers and scientists have not agreed on a definition of science, Popper’s theory is generally the preferred method. Empirical evidence, usually obtained by observational studies, is utilised when falsification testing is not possible (Okasha, 2016). Using these methods, science may be defined as the attempt to investigate and understand the physical world through reliable and testable observation. This definition of science implies that its sole focus revolves around getting closer to truth by rejecting what is false, which can be reinforced by reproducible and more accurate results. However, the reliability of scientific knowledge is only as dependable as the observer’s perception and the accuracy of the interpretation of the data they collect. Our perception of the world is relative to the capabilities of the human senses and subject to error (Lupyan, 2017), and science can also be misinterpreted by misinformed scientists, journalists and the public.

Truth and the scientific method are hard to define. Instead of waiting for old white men philosophising in armchairs to figure out what truth is we just test for things that aren’t true. Meaning nothing is ever scientifically “proven” to be true. We just get closer to the truth with higher accuracy over time. As an example, I’m pretty sure that if I pat a cat it won’t spontaneously combust. However, some day a little fuzz ball might explode immediately after being contacted and I’ll have no logical explanation for why. This doesn’t mean that the flames were magical, or more importantly, that I’m going to stop petting cats from here on out (that would be foolish to deprive myself of such a privilege). What I can say though, is that a cat is not a dog, because they are human made definitions that catagorise our buddies more easily. In saying that, since that catagorisation is based on our observation, rather than being universally true,we are restricted by our own senses and personal and collective biases.

The purpose of science is to create an improved overall understanding of the physical world, by observing predictable relationships that occur in constructed representative models (Gilbert, 1991). A model allows a scientist to make repeatable observations, without needing to wait for natural phenomena to occur and to test for a correlation in a specific environment. It also allows the observer to measure data with tools that exceed our human sensory capacity. Without controlling for variables within a scientific experiment, there is the chance that results may be altered by unseen interactions. Karl Popper believed that if an experiment cannot be tested for falsifiability, such as Freud’s psychoanalytic theory, it is to be considered pseudoscience (Harlow, Cummings, & Aberasturi, 2007). Popper defended this claim by stating that without the ability to control and measure the data, the experimenter is free to make non-objective and reductive claims based on their observations. Isolating variables in an experimental setting became the gold standard and scientists began to consider an experiment a method of falsification rather than a way of discovering what is true. By holding all the dependent variables in an experiment consistent, the independent variable can be changed by the scientist, making it possible to measure its effects more accurately (Peterson, Homer, & Wonderlich, 1982). Repeated testing can then be conducted to check for accuracy and reliability of the previous results. When being compared to testimony or observational knowledge, scientific knowledge contains much more reproducible results, as the data is derived from stringent and controlled testing that may be subjected to further investigation.

Science helps us to collectively predict life with better accuracy so we don’t have to try everything ourselves (if you’ve played the game Lemmings you’ll know why this is a good thing). It also helps us to isolate reasons behind why things happen. If a cat scratched your cousin’s face you could say the cat’s existence was the cause of the impromptu rhinoplasty. Alternatively, it may be due to them smothering it with their face. Science can help to test this in a safe environment, which can be repeated and altered for improved statistical significance(the ethics board probably wouldn’t like this one though). It can’t be confirmed that the face attack was the only cause, but the chances are pretty high. By retesting with different variables, such as using different limbs, not feeding the cat, or scaring it with a dog first you can isolate what is likely to be the cause of its fury. Leading to the conclusion that you probably shouldn’t headbutt or prod a kitty lest you bleed out slowly. While we probably wouldn’t need to carry out this test, we are more likely to believe the scientist’s rigorous assessment over our psychopathic young cousin’s, who has a history of swinging the cat by its tail. The difference is that the scientist can isolate variables, repeat the experiment for greater accuracy and reliability, and will generally not come to the unmovable conclusion that the cat is a big meanie.

The scientific methods used to conduct experiments are relatively robust when compared to purely observational studies, but human bias and misinterpretation of data can alter the reliability of an experiment. Amrhein et al. (2019) recently challenged the reliability of current statistical methods employed by scientists. They claimed that studies become redundant and negatively contribute to the scientific body of literature when scientists claim statistically significant results, or alternatively no results at all, when their P value is too close to 0.05. Relying on statistical significance can lead to the misinterpretation of data by scientists, which then may be used as supporting evidence for further research (Schmidt, & Rothman, 2014). Media miscommunication and the misinterpretation by the public are also ways that false conclusions can be drawn, and seemingly reliable results may appear to be false (Bell, 2016). Scientific knowledge itself is considered reliable when the human interaction with the experiment and other external variables are controlled, but the results become questionable when inaccurate statistics are concluded, an uneducated or generalised conclusion is drawn, or the reader of the work misinterprets the research. Therefore, it is hard to distinguish the reliability of scientific knowledge when the numerous interactions the experimenter has with the experiment and the misinformed interpretations of data can alter the results unknowingly.

Humans aren’t robots, so they’ll often get things wrong even if they don’t realise it. Even when we use statistics to confirm a hypothesis our human interpretation can get it wrong. To put it simply, we are not very reliable. In fact, our opinions will change from minute to minute, depending on how we feel. If you call your wife’s mother a manatee she will probably be less likely to believe anything you say, even if you have evidence that she did not in fact deliberately ask you to pick up the kids from school. Miscommunication extends beyond the household too. The role of a news outlet is to pass on information about the world in an interesting fashion. If they don’t, you won’t read it, so they fill it with clickbait and public recollections of the event. So, even if they aim to make the information as accurate as possible they will miss crucial details or an emotional retelling will direct the way you feel about it. By now we’ve all heard of the term “fake news”. In reality, all news is somewhat fake with the storyteller’s opinion weaved in. If you choose to watch or read from that source you probably already have an element of trust in them and will be more likely to believe them regardless. Science tries to eliminate these biases, but sometimes they slip through the cracks. Especially when the outcome of the experiment stands to benefit the person who funded the study. To summarise, humans make dumb mistakes, we don’t question the information as often if we like the source, and science is more reliable but not immune.

Justifying our trust in the understanding of the physical world requires the belief that our sensory and perceptual experiences are accurate (BonJour, 2007). Claude Bernard (1957), said that all our collective knowledge originated from observations of the world and the attempt to understand those observations. This suggests that the foundation of all our scientific discoveries are derived from experience limited by human perception. We routinely undergo hypothesis testing, with just our senses, and draw conclusions based on the results of such tests. Whether carrying out a scientific experiment or navigating the world in real time, we use perceptual experiences to provide justification for our beliefs (Siegel, 2013). We even developed tools that can measure and manipulate data beyond our perception, such as microscopes to see small objects, so that we can make can more accurate observations. But in the end, we still use our senses to understand the enhanced information from these tools, so the equipment is only as good as the interpretation of the user. This suggests that scientific knowledge is inherently entwined with perception but does not necessarily imply that perception is considered scientific knowledge. Scientific philosophers have attempted to create a boundary around what is considered science and what is not, to prevent justification of beliefs around perceived truths without evidence (Gieryn, 1983). However, these boundaries are constantly changing, making it challenging to establish a single definition of science. The reliability of scientific knowledge then becomes uncertain, as we rely on our methods being reproducible without being affected by the vulnerability of human error.

To live a somewhat normal and relatively stress free life we trust our senses to be correct, even if we know there is more to reality than what we can see. Evolution only does enough to make sure you don’t die. It doesn’t keep trending in a positive direction, unless more people with a certain trait reproduce than the amount that die without the trait. As an example, an antelope with stumpy legs may break its ankles less but a lion is going to catch that little horn nugget before it can snatch the gangly ones. What I’m trying to say is that we can’t see all colours, feel everything, hear all sounds, or understand half of what is going on around us. That would be exhausting for a body to maintain and arguably would not be worth the energy. However, we have built things like microscopes and hearing aids to improve our faulty abilities. The problem with being energy efficient meat bulbs is that our scientific understanding is completely built around our senses. So, that means they can’t be considered to be the all encompassing truth. It’s just the best we can do with what we’ve got at this point in time.

The purpose of perception is to use a combination of sensory apparatuses, such as eyes and ears, to transduce sensory data from the immediate environment and to estimate, adapt, and respond to this data accordingly (Lupyan, 2015). Hoffman (2009), argued that perception is not a true representation of the world and that evolution has provided a species-specific sensory framework that provides its user with enough information to navigate the world effectively. A reliable test is generally assumed to represent repeatable information when tested, but that does not necessarily mean that the data must be accurate. A clock may consistently be ten minutes faster than the correct time, which makes it reliable but not accurate. Cognitive hierarchical competition has also been seen to affect the reliability of perception (Lupyan, 2017). The results from a study by De Araujo et al. (2005), which labelled an odour either as cheddar cheese or body odour, found that participants ranked the smell labelled body odour worse than the one labelled cheddar cheese. This showed that the perceived opinion of a stimulus may be altered by competing senses in the form of a cognitive hierarchy. Senses, such as hearing and olfaction, have also been shown to decline with age, making in-between and within-subject design consistency less accurate (Wysocki, & Gilbert, 1989). Perception is only as accurate as the individual’s ability to sense, transduce, and judge the importance of information in a cognitively assessed hierarchy, but it is usually reliable to an individual subject given their senses do not deteriorate during the measurement process.

Sense are subjective to the person describing them and are not even directly translated into our brain as real information. As an example, sound gets converted from a specific frequency of air pressure, then into a physical movement in the ear, which is turned into an electric signal, and is then finally mapped onto the brain (it’s a little more complicated than this but we’ll stick with this for now). Then that is interpreted into an idea or pattern that you compare to other sounds to summarise that sound as your cat meowing. Over time, just like everything in our body, our senses degrade and we struggle to pick up on what might have been easier when we were young. Also, our previous experiences shape what we predict things to be and we make comparisons when other senses are detected in close succession. One day you decide your shirt is blue but the next day you compare it to your friend’s blue shirt and decide it is actually more of a teal colour.

Science provides a tool for information to be measured with defined reproducibility rules. The objective of science is to create an unbiased worldview and to falsify perceived information to get closer towards an objectively truer understanding of the environment. Adherence to strict scientific methods should result in higher reliability and more accurate observations of the world. If the statistics of these methods is not interpreted correctly, the reliability of research becomes questionable and may contribute towards compounding inaccurate conclusions in the scientific body of literature. The reliability of scientific knowledge is also affected by the inconsistencies of our individual and comparative perception. Our experimental methods are limited by our current understanding and our ability to transduce sensory information. We can measure information beyond our human capabilities to a certain extent, but they must be converted to a format that we can comprehend. In conclusion, scientific methods are currently the most reliable way we have of acquiring knowledge, but at this point it is unclear whether scientific knowledge is the most reliable knowledge we have without a way of confirming the reliability of our senses and finding a more accurate way to interpret data.

Science is good but not perfect, humans save energy by making things up in their brains, this means our scientific understanding is restricted by our perception, nothing is real, everything is subjective, illuminati illuminati, the government are lizards. Have a good day.

Amrhein, V., Greenland, S., & McShane, B. (2019, March). Scientists rise up against statistical significance. Nature, 567, 305–307. doi: 10.1038/d41586–019–00857–9

Bell, A. (2016). Media (mis) communication on the science of climate change. Public understanding of science.

Bernard, C. (1957). An introduction to the study of experimental medicine (Vol. 400). Courier Corporation.

BonJour, L. (2007). Epistemological problems of perception.

De Araujo, I. E., Rolls, E. T., Velazco, M. I., Margot, C., & Cayeux, I. (2005). Cognitive modulation of olfactory processing. Neuron, 46(4), 671–679.

Gieryn, T. F. (1983). Boundary-work and the demarcation of science from non-science: Strains and interests in professional ideologies of scientists. American sociological review, 781–795.

Gilbert, S. W. (1991). Model building and a definition of science. Journal of Research in Science Teaching, 28(1), 73–79.

Harlow, S., Cummings, R., & Aberasturi, S. M. (2007, March). Karl Popper and Jean Piaget: A rationale for constructivism. In The Educational Forum (Vol. 71, №1, pp. 41–48). Taylor & Francis Group.

Hoffman, D. D. (2009). The user-interface theory of perception: Natural selection drives true perception to swift extinction. In In.

Lupyan, G. (2015). Cognitive penetrability of perception in the age of prediction: Predictive systems are penetrable systems. Review of philosophy and psychology, 6(4), 547–569.

Lupyan, G. (2017). How reliable is perception?. philosophical topics, 45(1), 81–106.

Okasha, S. (2016). Philosophy of Science: Very Short Introduction. Oxford University Press.

Peterson, L., Homer, A. L., & Wonderlich, S. A. (1982). The integrity of independent variables in behavior analysis. Journal of applied behavior analysis, 15(4), 477–492.

Schmidt, M., & Rothman, K. J. (2014). Mistaken inference caused by reliance on and misinterpretation of a significance test. International journal of cardiology, 177(3), 1089–1090.

Siegel, S. (2013). The epistemic impact of the etiology of experience. Philosophical Studies, 162(3), 697–722.

Wysocki, C. J., & Gilbert, A. N. (1989). National Geographic Smell Survey: effects of age are heterogenous. Annals of the New York Academy of Sciences, 561(1), 12–28.