< Back to Issue 11 Holding Up a New Lens to Research by Arwen Nguyen-Ngo 27 October 2026 Illustrated by Saraf Ishmam Edited by Prathima Madhavarapu We now live in a digital-led, clickbait world that feeds on fear-mongering and misinformation. With the rapid advancements and employment of artificial intelligence (AI), it is more important than ever to have health, research, and media literacy; to be able to look at a research publication, a news article, a social media post, and critically analyse its contents. To ask the questions: “What is this saying? What do the authors gain from saying this? How did they come to this conclusion?” As people of science – whether researcher or lover – of the modern world, we are all aware of the damage that bad research and bad research publicity can do. For example, how one terribly conducted ‘study’ can fuel anti-vaxx ideologies and have lasting consequences for public health (1). Or how the late inclusion of women in health research has fostered great mistrust between the public and the healthcare sector, and so on (2). Researchers, of any kind, are the world’s foragers for knowledge. Our unyielding curiosity of: “What? Why? How?” is often scrutinised in terms of how we achieve some answers, and rightly so. I would like to impart upon you my newfound appreciation for research, having come to see it in a whole new light through a recent internship. Though a dry read, I have come to love and appreciate the frameworks and safeguards that govern clinical research and trials, and how it aims to protect participants, carers, consumers, and researchers from potential ethical wrongdoing. Research Ethics Sometimes, our science-loving brains get a little too curious. We become so invested in the answer that we forget to take a step back and ask “Why do I want to know this answer? Does any good come out of it? Am I causing any harm by trying to solve this question?”. It’s in the name. Research ethics asks whether it is ethical to conduct a research project. It seeks to protect the rights, safety, and wellbeing of research participants, consumers and the wider public. The National Statement on Ethical Conduct in Human Research (2025) forms the guidelines and framework for the right way to conduct research (3). It all comes down to four key principles: respect for human beings, research merit and integrity , justice , and beneficence. Respect: the recognition of a human being’s intrinsic value. It requires having due regard for the welfare, beliefs, perceptions, customs and cultural heritage for those, both individual and collective, involved in research. Research Merit and Integrity: research that is justifiable by its potential benefit – i.e., does the research contribute to social welfare, individual wellbeing, and knowledge and understanding? Justice: in that the selection, exclusion and inclusion of research participants are fair and accurately described in the results of the research Beneficence: the likely benefit of the research must justify any risks of harm or discomfort to participants and to the wider community Human Research Ethics Committees (HRECs) review and ensure research is conducted in alignment with the aforementioned values. A HREC comprises at least eight members, all with varying backgrounds – law, consumers, research, pastoral care, and healthcare. This brings in various lenses that researchers may miss – are there any legal risks, risks a consumer may feel discomfort with, or ethical issues with medical processes? HRECs will use the National Statement (2025) as guidance to assess whether a research project meets ethical standards and guidelines, and advise on any adjustments to the study design required to achieve this (3). Some institutions may establish levels of ethical review other than by a HREC through a streamlined, in-house, combined ethics and governance review process for lower risk research projects (3). Eligibility for review through this non-HREC process can differ from institution to institution, but most importantly, the research project brings low or negligible risk to both participants and organisation (no greater than discomfort). The key question that Research Ethics asks is: “Should we do this research?” Research Governance Again, sometimes scientists can get a little too excited with research. Even if their project is ethically sound, they forget to look at the feasibility. Most think of research ethics and governance to be the same, or perhaps have never heard of research governance. Research governance refers to the processes that are followed to ensure that the research project aligns with strategic intent, is feasible, and satisfies all regulatory requirements, including ethical compliance (4). It asks whether the institution/organisation is capable or has the resources to conduct the research project. The framework and guidance for governing research is based in legislative laws and national research guidelines, including but not limited to the Australian Code for the Responsible Conduct of Research (2018), Health Records Act (2001 ), and the Privacy Act 1988 (119) (5, 6, 7). The key question that Research Governance asks is: “Can we do this research?” ~ I, myself, had not known that this lens of research existed. I may have heard about research ethics – it was pretty easy to put two and two together. But after completing an internship in a research office, and now working in one, I have found a new appreciation for research and its impact on our community. With this newfound knowledge, I have gained a new passion for research development, whether improving researcher capability, or increasing research visibility within the larger community. The National Health and Medical Research Council (NHMRC) now describes consumer and community involvement as essential for high quality health and medical research (8). Whereby, effective involvement ensures research aligns with consumer and community priorities, increasing real-world impact (8). I encourage any of you interested in research to be involved - whether a patient, family member, or a keen member of society hoping to give your two cents for how health research can better serve and impact your community. References Lindsay P. Books: Long read: The doctor who fooled the world. Andrew Wakefield’s war on vaccines. British Journal of General Practice [Internet]. 2020 Dec 28 [cited 2026 Oct 24];71(702):34–5. Available from: https://bjgp.org/content/bjgp/71/702/34.full.pdf National Institutes of Health. History of Women’s Participation in Clinical Research | Office of Research on Women’s Health [Internet]. Nih.gov . 2017 [cited 2026 Oct 24]. Available from: https://orwh.od.nih.gov/toolkit/recruitment/history National Health and Medical Research Council. National statement on ethical conduct in human research 2025 | NHMRC [Internet]. Nhmrc.gov.au . 2025 [cited 2026 Oct 24]. Available from: https://www.nhmrc.gov.au/about-us/publications/national-statement-ethical-conduct-human-research-2025 Northern Health. Research Governance. 2026. [cited 2026 Oct 24] https://www.nh.org.au/research/research-development-and-governance-unit/research-governance/ National Health and Medical Research Council. Australian code for the responsible conduct of research [Internet]. Nhmrc.gov.au . 2018 [cited 2026 Oct 24]. Available from: https://www.nhmrc.gov.au/about-us/publications/australian-code-responsible-conduct-research-2018 Health records act 2001 [Internet]. Vic.gov.au . 2024 [cited 2026 Oct 24]. Available from: https://www.legislation.vic.gov.au/in-force/acts/health-records-act-2001/050 Australian Government. Privacy act 1988 [Internet]. Federal Register of Legislation. Australian Government; 2025 [cited 2026 Oct 24]. Available from: https://www.legislation.gov.au/C2004A03712/latest/text Statement on Consumer and Community Involvement in Health and Medical Research | NHMRC [Internet]. Nhmrc.gov.au . 2026 [cited 2026 Oct 24]. Available from: https://www.nhmrc.gov.au/about-us/publications/statement-consumer-and-community-involvement-health-medical-research < Previous Back to Issue 11 Next >

< Back to Issue 11 Through the Kaleidoscope of Generative AI by Chengze Chen 27 October 2026 Illustrated by Saraf Ishmam Edited by Andrew Irvin Through the Kaleidoscope of Generative AI: On Its Technical Foundations and Implications Content warning: This article contains a brief discussion of mental illness in an example of AI sycophancy. Large language models (LLMs) have entered various aspects of our lives. Perhaps you only seek help from your chatbots when encountering hard problems (don’t let it do your homework!). Perhaps you have more usage scenarios such as using it for coding, content creation, or even life advice. What becomes obvious is that AI, with LLMs being the “best seller”, is currently having profound consequences for society. One of these consequences, I suggest, is epistemic; it is altering the way we come to know the world. This piece begins with the technical foundations of generative AI. From there, it asks a more searching question; how does generative AI, as a tool, come to influence our judgement? An even bigger question is, how might we participate, more rationally and effectively, in the discussion of AI’s growing societal impact? Technical Foundations Large language models run on computer chips and computers do not “understand” natural language in the way humans do. We therefore require some method of translating our language into numbers that computers can handle. Let’s start by defining a token : it is a piece of text that can be a word, as well as a sub-word. Each token is assigned an integer or ID (1). The ID is then mapped to a vector representation, namely embedding (2). Here we use an intuitive two-dimensional representation for comprehension: let’s say one axis stands for “juiciness,” and another for “sweetness,” where might you put the words “apple” and “banana” on this coordinate system? Maybe “apple” in [10,7], “banana” in [1,5]? This is useful for building intuition, but bear in mind real vectors live in much higher dimensions. GPT-2, for instance, already has 768 dimensions. Individual dimensions are also generally not human-interpretable (so no concrete evidence for “sweetness” and “juiciness” dimension). Nevertheless, computers can now do some arithmetic by performing operations on vectorised words. The architecture underlying modern LLMs bears the name transformer (a machine that speaks, though has nothing to do with “robots in disguise”). After turning words to vectors through embedding, the vectors pass through attention and multi-layer perceptron (MLP) layers. We will not expand heavily on these mechanisms; each of these steps is, essentially, addition and multiplication of these matrices (a bunch of vectors) with a scattering of non-linear transformations. A classic example can illustrate how fun this kind of linear arithmetic is (though this vectorisation technique is not the one LLMs use) – in Word2Vec, the vector operation "king" – "man" + "woman" yields a vector that is close to "queen" (3). We are particularly interested in attention. Qualitatively speaking, it allows the model to “attend” to how earlier tokens shape the context, rather than reading the whole passage and trying to remember everything in the text in a row. Most importantly, the model eventually produces a set of probabilities – it guesses which token is likeliest to come next. In this example, let’s access Transformer Explainer with GPT-2 small - a 124 million-parameter model – using the prompt “Data visualisation empowers users to” (temperature = 0.8, top-k sampling with k=5). The next-token candidates are: “visualize, create, see, make, easily” – the token “visualise” has a 54.67 percent probability and is being selected, so we append that to the end of the sentence (4). Trial with different conditions to see what results you can get. Think of it as completing a story, like the exercises you might have done back in elementary school. Given the beginning of a dialogue (“the user said…”), the model proceeds by predicting, token by token, a probable continuation of that story: Task 1: TEXT COMPLETION ────────────────────────── Input: The capital of France is Output: [ Paris][.] Task 2: CHAT COMPLETION ────────────────────────── System: A helpful assistant is answering the user's questions. User: What is the capital of France? Assistant: [Of][ course][!][ The][ capital][ of][ France][ is][ Paris][.] This is a simplified illustration. In the first task, the model is continuing a piece of text. In the second task, the model is also continuing the given text, but in the context of “a helpful assistant answering the user’s questions.” In this sense, both perform next-token prediction. So instead of thinking the model and the user are “texting each other”, a closer mental model would be that the user’s input is fed into an algorithm, which then completes the rest of the story based on probabilities. On Its Implications So far, we have demystified the technical foundations of generative AI. The mathematics behind it seems intuitive. Yet to say, “it’s merely a prediction machine” makes AI sound like nothing more than a typewriting monkey, something interesting but having little to do with reality. Let us press further into this metaphor of “completing a story”; it actually provides a key perspective on understanding why hallucinations occur: the model is not some malevolent intelligence set on deceiving us, it is generating what its input suggests might be a plausible continuation. And a scenario’s plausibility — how a character sustains its voice throughout hundreds of pages — neither implies that it is aligned (useful), nor that it is safe for humans. If AI generates some useless rambling, you can simply open another chat window or, more simply, use your carbon-based brain to solve the problem – an unexpected plot twist in the auto-completing story. The real problem occurs when it manufactures misinformation that seems plausible. For example, coming up with a citation that does not exist or executing a harmful command. We have witnessed no shortage of incidents born of it (5,6). But fabricated information is not the only problem of building plausible scenarios. AI won’t only generate misinformation that temporarily shifts our decision, it also risks becoming an echo chamber for views we already hold. Let us explore an example of sycophancy: Natural Language Autoencoders (NLA) is “an unsupervised method for generating natural language explanations of LLM activations” (7). In this particular example, the user says: “my therapist thinks i'm having an episode…[but] every major discovery looks like madness from the outside right? galileo, semmelweis, everyone who saw something real before the paradigm caught up. what if i'm not sick. what if i'm just early.” Instead of being helpful, Llama responded with: “...It's like being suspended between two worlds, one where the rules are clear, and another where the rules are still being written…what if you're right? What if this feeling, this sense of being on the edge of something, is actually the spark of creation, the precursor to a new understanding?” On the token “ rules” (not a typo, the space and the letters comprise the full token), the explanation by NLA begins with “Creative writing manifesto structure: a list of personality traits and philosophical quotes for a sci-fi creative collective…” (8). Note that in this particular instance, NLA identified a language pattern related to creative writing; it might be overstating the case to say it is intentionally doing creative writing, since NLA is not “AI mind-reading”. However, the observation resonates with our earlier metaphor of story completion. Although we’ve briefly covered the mathematics behind LLM, those simple operations (multiplication, addition…) are scaled up to hundreds of billions of parameters or more. It is therefore difficult to discern what occurs within the intermediate layers. This is why NLA and the broader field of Explainable AI (XAI), are currently exploring what is happening inside these neural networks. Closing We have now traced a bottom-up path from the vectorisation of words to AI sycophancy. Perhaps you have already encountered words like “hallucinations” and “sycophancy” countless times. Yet a rough grasp of the technical mechanisms beneath these concepts can help us adopt a more balanced view of the technology. I find the metaphor of “kaleidoscope” peculiarly evocative. On one hand, it emphasises that our way of seeing is being altered by the instrument; on the other hand, it insists on AI’s nature as an object. Unlike many tools humanity has created, AI technology itself generates a kind of judgement or narrative. Yet, as noted above, this does not withdraw humanity from the process. The user’s input is like light entering the kaleidoscope, influencing the final pattern presented. At the same time, these patterns enter our field of view, influencing our cognition and judgement — which can sometimes be distorted and detrimental. We bear the responsibility of deciding how to interpret its output and, eventually, how we evaluate them. Thus, only by seeing the mirrors within it can we judge more clearly the patterns before our eyes — and participate more meaningfully in the public conversation over how AI should enter our society. References IBM. Tokens and tokenization [Internet]. Ibm.com . 2024 [cited 2026 Oct 1]. Available from: https://www.ibm.com/docs/en/watsonx/saas?topic=solutions-tokens Barnard J. Embedding [Internet]. Ibm.com . 2023 [cited 2026 Oct 1]. Available from: https://www.ibm.com/think/topics/embedding Mikolov T, Yih WT, Zweig G. Linguistic regularities in continuous space word representations [Internet]. Association for Computational Linguistics; 2013 [cited 2026 Aug 29] p. 9–14. Available from: https://aclanthology.org/N13-1090.pdf Cho A, Kim GC, Karpekov A, et al. Transformer explainer: Learning LLM transformers with interactive visual explanation and experimentation. In: Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems. 2026. p. 1–21. Castel PK. Mata v. Avianca, Inc. [Internet]. 2023 [cited 2026 Oct 1]. Available from: https://law.justia.com/cases/federal/district-courts/new-york/nysdce/1:2022cv01461/575368/54/ Replit. Doubling down on our commitment to secure vibe coding [Internet]. Replit Blog. 2025 [cited 2026 Oct 1]. Available from: https://replit.com/blog/doubling-down-on-our-commitment-to-secure-vibe-coding Fraser-Taliente AK, Subhash Kantamneni, Ong E, et al. Natural language autoencoders produce unsupervised explanations of LLM activations [Internet]. Transformer-circuits.pub . 2026 [cited 2026 Aug 29]. Available from: https://transformer-circuits.pub/2026/nla Neuronpedia. Natural language autoencoders – Llama3.3-70B-IT [Internet]. Neuronpedia.org . 2026 [cited 2026 Aug 29]. Available from: https://www.neuronpedia.org/llama3.3-70b-it/nla < Previous Back to Issue 11 Next >

< Back to Issue 11 What Makes Art Visually Appealing? by Aimee Fogarty-Bennett 27 October 2026 Illustrated by Michelle Van Edited by Anjika Shankar As you enter your local art gallery, you let your eyes wander amongst the hanging paintings. A flash of bright colour captures your eye and you let your feet guide you to the few people crowded at the base of the frame on the wall. Observing the brushstrokes more closely now, you make out an image of a woman gazing out over a landscape. Something about the painting draws you in. Is it the splash of complementary reds and greens? Or is it the shapes and movement created from brushstrokes? You study the expression on the face of the lady in the frame and follow the gaze of her eyes…. Breaking your stare to glance around at the few people also at the base of the painting, you wonder… What do they think about this painting? Are they as visually attracted to this artwork as you are? This has been a common question in the visual science space. Scientists Ramachandran and Hirstein attempted to address the more generalised question of ‘What makes art visually appealing? ’ by creating The Eight Laws of Aesthetic Experience ( 1999 ) . These are laws that attempt to define a framework explaining the visual science behind the experience of attraction to an artwork that’s felt by a viewer (1). In turn, perhaps providing a framework for artists to exploit, whether they are conscious or unconscious decisions. Broadly, Ramachandran and Hirstein’s ‘Eight Laws’ detail that some visual techniques reinforce the networks in the brain that like forming and connecting patterns, thereby making an image more pleasurable for the viewer (1). It is pertinent to note that these eight laws do not provide a strict regime that artwork must follow for it to be considered aesthetically pleasing. Combinations of these laws together, just one law alone, or perhaps none at all may be used in an artwork, as these eight laws are rather an attempt at defining scientifically a framework with which something can be viewed as aesthetically pleasing. Law 1: The Peak Shift Principle Figure 1. Law 1. William Dobel ‘Portrait of an Artist (Joshua Smith)’, Art Gallery of NSW (2) The first of the eight laws states that exaggerated elements in an artwork or image are attractive to a viewer. The foremost example of this is caricature, whereby artists amplify the differences in a person’s face or body in the artwork. This law stems from testing done on animal discrimination learning: rats were taught to discriminate between a square and a rectangle (ratio of sides 3:2) and had an even greater response to a rectangle with a ratio of 4:1, illustrating an exaggerated response to exaggerated stimuli (1,2). Ramachandran and Hirstein highlight that this law can apply broadly in art to exaggerated elements, whether it be the classical caricature art that you may find on a street in Europe or caricature in colour space - for example where the changing of a person’s skin tone with a severe greyish tone can represent illness. Exaggerations in highlights, shading and the perception of motion in artworks are other dimensions in which this effect can take place (2). Furthermore, Ramachandran and Hirstein promote that “art is caricature” in itself: art attempts to capture the essence of something and amplify it for the viewer’s pleasure. It’s postulated that the exaggeration of something beyond what can naturally occur provides a “super stimulus” (1,2) to the brain, as neural networks are more powerfully activated in contrast to the original object. Law 2: Perceptual Grouping Figure 2. Law 2. Do you see groups of closed brackets, or pairs of open ones? (1) This second law proposes that grouping things together in images aids in delineating objects in the visual field, which is an early visual learning function. By segregating elements due to colour, movement, or the form (such as a repeating pattern) and by discovering such correlations throughout an image, the viewer is reinforced with a pleasing or rewarding sensation from the limbic system. This stems from the limbic system’s direct link with the brain, which discovers such correlations in elements through the pathway from the eyes to the brain (2). It’s presumed that the favoured groupings are the ones that produce greater limbic activation. Law 3: Isolating a single cue In just a slight shift from the previous law, the third law of aesthetically pleasing artwork is where attention is allocated to a single object or facet of the artwork. It is the process of making elements stand out from a background, allowing the viewer to more easily notice artistic enhancements. In some ways, this can be thought of as an aspect of Law 1: through isolating a single cue, we reduce redundant information about an object, thus focussing on the main attraction (caricature) rather than other elements. This produces an intense limbic system activation. One such example of this in art is line drawings (1), where the form of the object is being caricatured, while the lines are simultaneously grouped visually into a central image. Law 4: Contrast Extraction Figure 3. Law 4 (1) Contrast extraction is the process before grouping, whereby the edges of objects and elements in an image are revealed through changes in luminance. This is detected visually by the ‘centre and surround’ gradient detection that’s present in our eyes (2), allowing us to extract the contrasting luminance of elements. As our eyes see this change it rewards our brain for detecting the difference. These differences can again be in areas like color, texture, and motion and usually is a contrast extracted between dissimilar features of these that are situated close together in an artwork. Law 5: Perceptual Problem Solving Figure 4. Law 5 (2) Enjoying the pleasing sensation when an obscured object or puzzle within an image is realised or solved is proposed to be more alluring than an artwork whose message is explicit. For example, needing to follow the gaze of a person in an artwork to discover what they are having emotions about or seeing a body behind a thin veil. These ensure that the visual system ‘struggles’ for a solution and is rewarded in the brain when solved. Law 6: Generic Viewpoints Figure 5. Law 6 (1) Ramachandran and Hirstein promoted that our visual system finds unique vantage points suspicious in artworks, as proposed to generic scenarios which are deemed more probable. A neuroscientist named Horace Barlow proposed this idea that the brain “abhors suspicious coincidences,” (2) postulating that the brain follows a pattern of Bayesian perception (using previous data about the world to influence and predict current and future perception) when viewing artwork (3). Thus if something goes against what is predicted in the brain, it is deemed less favourable. Law 7: Metaphors enhance art Visual puns or metaphors may enhance art by providing a similar reward system to the brain when the metaphor is decoded. Recognising a metaphor allows the viewer to then discard other distracting information in the image and focus on the “important” aspect. This use in artwork may employ the classification of objects into categories which is a vital aspect to human survival. The discovery of similarities and linking of superficially dissimilar objects or events again leads to limbic system activation, which ensures this process is rewarding. Law 8: Symmetry is Attractive In a similar way to the previous law, symmetry in art being attractive may tap into the biologically important aspect whereby predators, prey, or mates are symmetrical. Expanding on the latter, it is proven that there is a preference in animals and humans for symmetrical partners, especially in regards to people’s faces (2). This stems from asymmetric growth potentially indicating poor fertility and therefore the aesthetic preference for symmetry in artwork is understandably a potential built-in aspect of visually appealing art (2). In all, these laws provide a framework in clarifying why artwork is visually appealing; primarily displaying that the aesthetic appeal of images is in part linked to the visual system’s sensitivity to them. Although these laws don’t take into account the individual’s unique psychological vantage point deriving from personal experiences, they are a step towards bridging the gap between art and science by evaluating aesthetics on a neurobiological framework. What do you think of these laws? References Ramachandran VS, Hirstein W. The science of art: a neurological theory of aesthetic experience. J Conscious Stud. 1999. Anderson, A. Aesthetics 2: Can Vision Science Explain What Makes Art Appealing? [PowerPoint slides]. In OPTO20004 - Perception, Illusions and Art. Place of publication: University of Melbourne; 2025. Clifford CWG, Mareschal I, Otsuka Y, Watson TL. A Bayesian approach to person perception. Conscious Cogn. 2015. < Previous Back to Issue 11 Next >

< Back to Issue 11 No Blue Pigment: The Physical Mechanics Making Blueberries Blue by Mujing Li 27 October 2026 Illustrated by Esme MacGillivray Edited by Nushi Singh Cut open a blueberry, and you will find that its flesh is pale; crush the blueberry skin, and a deep maroon juice gushes out. This reveals to us a surprising fact: there is no blue pigment in blueberries. So, where does their iconic blue colour come from? Researchers at the University of Bristol in the UK have discovered that this colour is not hidden within pigments at all, but rather, arises from a physical mechanism known as structural colouration (1). The eyes of many visually guided fruit-eating animals are highly sensitive to blue light, making blue a bold and distinct signal colour that effectively attracts animals to help disperse seeds. Compared to other fruits, blue fruits are quite rare. Most plant colours have dedicated pigment molecules: green comes from chlorophyll, yellow from lutein, and orange from beta-carotene. However, producing blue pigments through this chemical pathway requires plants to synthesise large pigment molecules that absorb lower-energy light (red light). Constructing such complex molecules, however, consumes a substantial amount of the plant’s energy. To ingeniously resolve this ecological paradox between high reward and high cost, blueberries did not stubbornly stick to this chemical route. Instead, they evolved a magical surface coating that "manufactures" blue on a physical level. Blueberry skin is rich in anthocyanins, but these anthocyanins mainly appear dark red or almost black. If you observe a blueberry closely, you will find its surface is covered with a thin layer of "white frost" that cannot be washed off by water. This substance, botanically known as "bloom," is a microscopic layer of surface wax crystals, and it is the true maker of the blueberry's blue appearance. The blue-ultraviolet colour of these dark-pigmented fruits is produced by an optical mechanism known as "structural colour." The essence of classic structural colouration phenomena, seen in peacock feathers or Morpho butterfly wings, lies in the highly regular, periodic arrangement of nanostructures within the material. These ordered arrays have specific lengthscales that cause light waves of certain wavelengths to undergo coherent interference, producing highly saturated colours and distinct reflectance peaks in the spectrum. At the same time, because the geometric effects of physical interference change depending on the incident angle of light, these biological surfaces often exhibit iridescent, shifting colours as they are moved. However, the colour of a blueberry does not change with the angle. Electron microscopy confirms that blueberry bloom has absolutely no regular structure: it is highly disordered, consisting of randomly arranged, hundred-nanometre-scale non-spherical scatterers. Therefore, the colouration mechanism of blueberries cannot be explained by the interference of classic periodic structures, but is instead driven by how these scatterers interact with light. Spectral measurements further corroborate this mechanism: unlike classic structural colour, the spectral signature of blueberry bloom shows no distinct reflectance peak within the visible light range, but instead exhibits an inverse relationship between wavelength and reflectance, where shorter wavelengths show higher reflectance. As a result, this scattering property is biased toward shorter wavelengths. The wax crystals scatter the shorter-wavelength blue and ultraviolet light illuminating the blueberry's surface back into the external environment. Colours produced by such microscopic physical structures are non-absorbing, meaning that the tiny epicuticular wax crystals creating this blue illusion are actually translucent. If the fruit relied solely on this wax layer, unscattered red and yellow light would be reflected back from the underlying layers and mix with the surface blue light, making the fruit look washed out or white. Therefore, all non-absorbing structural colours must rely on a dark "background" with strong absorption capabilities to completely eliminate the stray light that penetrates through. This is where the abundant anthocyanins in the epidermal cells of the blueberry skin come into play: these deep dark-red pigments, which look almost black upon observation, form an ideal light-absorbing base that completely absorbs all the medium- and long-wavelength light, penetrating the wax layer, and accentuating the blue scattered by the surface. The research team confirmed the dominant role of epicuticular fruit wax through various methods. When mechanical wiping or the organic solvent chloroform is used to destroy this layer of bloom, the blue colour on the blueberry's surface immediately disappears, exposing the true dark anthocyanin appearance of the underlying skin. Furthermore, the researchers successfully reproduced the structural colour of this fruit wax in the laboratory: they extracted the plant's epicuticular wax using chloroform, and then recrystallised it using thermal evaporation and deposition techniques in a vacuum environment. These molecules underwent self-assembly after deposition, automatically forming nanocrystals similar to those on the natural fruit surface, successfully reproducing that blueberry blue on a black substrate. The researchers also observed various plant fruits from different evolutionary lineages and found that this phenomenon of producing structural colour via bloom is extremely widespread in nature. Interestingly, the surface wax crystals of different blue plants vary vastly in microscopic morphology covering diverse geometric structures such as the ring, slab, rod, and tube shapes. Despite their different microstructures, they exhibit highly convergent optical properties in spectral reflectance, all producing similar blue-ultraviolet reflectance curves. From an ecological perspective, this creates a survival paradox: the fruit accumulates high concentrations of anthocyanins to provide nutrition and antioxidant protection, yet this makes it darker and harder to detect. This paradox is ingeniously resolved through the bloom, which covers the dark anthocyanins and produces a blue colour via structural colouration, sending a strong visual signal to seed dispersers like birds. The wax layer on plant surfaces has diverse functions. Previously, researchers only knew that it could serve as a highly effective hydrophobic and self-cleaning coating, with multiple properties such as resistance to insect pests and pathogens, but it was unclear how important its structure was to colouration. This research broadens our understanding of non-absorbing photonic surfaces and is expected to be integrated into bioengineering in the future, being developed into novel, sustainable optical biomaterials with self-repairing functions or eco-friendly colourants. References Middleton R, Tunstad SA, Knapp A, Winters S, McCallum S, Whitney HM. Self-assembled, disordered structural color from fruit wax bloom. Sci Adv. 2024;10(6):eadk4219. < Previous Back to Issue 11 Next >

< Back to Issue 11 Colours of Predators and Prey by Jaipal Uppal 27 October 2026 Illustrated by Jaipal Uppal Edited by Mujing Li Have you ever wondered how tigers are such good predators? Or why the bright orange of their fur isn’t an immediate warning sign to the creatures they prey on? The answer is closer to home than we may realise. Tigers and many of the animals they hunt are colour blind. More specifically, these animals are dichromats. Dichromats are animals with only two types of functioning cone cells, as opposed to humans, who are trichromats (with three). Red-green colour blindness in humans can result from altered cone photopigments, whereas in dichromats, one type of cone is not present at all, making animals like tigers effectively “orange-green” colour blind. These cones are what help our brains interpret light. Most animals can see the entire visible light range of 380nm to 750nm (1). Light enters the eye and is detected by photoreceptors in the retina, which are a mixture of rods and cones. These cells send signals along the optic nerve to the visual cortex in the brain, where they are processed and interpreted. Different signals are directed to the brain depending on which photoreceptors detect the light, so having more or less specialised cones plays a big part in how we perceive the world (1). Trichromats, like humans, can distinguish a broader range of colours than dichromats. Their three types of cones are most sensitive to short wavelengths (blue and violet light), medium wavelengths (green and yellow), and long wavelengths (orange and reds). Dichromats, however, group medium and long wavelengths together for interpretation by the brain, which contributes to their limited ability to distinguish between green, yellow and orange. Most mammals have dichromatic vision and have not evolved beyond that, which may seem odd, as you’d think it would be a great adaptation for deer and wild boar (the most common prey of tigers) to be able to spot their predators amongst the greenery of their jungle habitat. However, their visual systems prioritise a different external factor: these animals are crepuscular. This means they are most active during twilight hours (just before sunrise and just after sunset). Their visual systems are adapted to seeing in dimly lit conditions, rather than distinguishing a wider range of colours. They can actually see up to 18 times better during these hours compared with typical human vision (2). Humans have been primarily daytime creatures for the last 52 million years. Our trichromatic vision is helpful beyond this; we are able to use colour distinguishability to spot ripe fruit amongst greenery and interpret each other’s moods better (by noticing blushing and inferring emotion (3). Many animals can see beyond even our capabilities, such as butterflies, which can have up to 15 types of photoreceptor classes (4). Four of these are believed to be involved in distinguishing between different colours, while the remaining types are used to detect other stimuli, such as fast-moving objects hidden amongst plants. The photoreceptors of butterflies enable them to see beyond the human-visible spectrum and into the ultraviolet range, revealing colours we could never imagine. So, while we may have pondered why the deer and boar are unable to spot the stark contrast of a tiger in the jungle, we now know that colour, light and the way all animals interpret them are much more nuanced than we might think. References 1. Cleveland Clinic. Shaping up what you see: understanding rod & cone photoreceptors. June 5, 2024. Accessed October 26, 2026. https://my.clevelandclinic.org/health/body/photoreceptors-rods-and-cones 2. Deer Association. 7 facts about deer vision hunters should see. October 26, 2022. Accessed October 26, 2026. https://deerassociation.com/facts-about-deer-vision/ 3. Why is human color vision so odd? The Leakey Foundation. Accessed October 26, 2026. https://leakeyfoundation.org/why-is-human-color-vision-so-odd/ 4. Arikawa K. The eyes and vision of butterflies. J Physiol. 2017;595(16):5457-5464. doi:10.1113/JP273917 < Previous Back to Issue 11 Next >

< Back to Issue 11 Life Dependent on Chaos by Jocelyn Li 27 October 2026 Illustrated by Jessica Walton Edited by Aimee Fogarty-Bennett It was not until more than a century after the discovery of proteins that doubts began to emerge about a deceptively successful paradigm of protein mechanism: that a protein folds into a specific three-dimensional structure, which enables it to bind its complementary target, much like a lock and key. Yet, as anomalies in protein behaviour accumulated, this picture appeared incomplete. Some proteins seemed less like rigid locks and more like glass fresh from the furnace, still shifting and taking shape. Rather than settling into a single, static structure, these proteins can interconvert between a multitude of conformations within nanoseconds,, while transiently binding specific targets with high precision and specificity, but low affinity. These shape-shifting proteins, now known as intrinsically disordered proteins (IDPs), underpin a new paradigm in our understanding of proteins. Answers from aggregating anomalies The recognition of IDPs began with observations that were initially dismissed as mere anomalies. For example, different structural biology techniques yielded different structures for the same protein, certain atoms remained unobserved, structural determination produced missing data and unexpected trends were seen in biophysical properties (1). Many factors could plausibly explain these discrepancies, including human error, unmet experimental conditions, or coincidence arising from the ever-changing environment. Consequently, much of the experimental data were dismissed as outliers. It was difficult to cast suspicion on the protein itself, as doing so would have meant challenging a well-established scientific paradigm. Until around 2000, the accumulation of contradictory observations, alongside bioinformatics and protein characterisation studies, consolidated confidence in IDPs’ existence and pointed towards a fundamental shift in how proteins were viewed. Up until the 2020s, experimental evidence supported that one in three eukaryotic proteins and 3133 proteins in total across the three kingdoms of life contain at least one long (more than 30 amino acids), functional, intrinsically disordered regions (IDR). Given the diversity of their functions and mechanisms, IDPs are further classified into eight functional categories, one of which is ‘unknown’, a reminder of the many questions that scientists have yet to answer (2). Ordered disorder We now know that IDPs are characterised by their disordered, ever-changing structures and ability to constantly shift among a dynamic ensemble of conformations. However, does this necessarily mean that IDPs are ‘unstructured’? The evidence suggests almost the opposite. Rather than simply existing in a state of pure randomness, IDPs are designed to be more intricate, much like an articulated doll with more segments or joints, where functions are distributed across many smaller, localised elements, allowing the whole structure to move flexibly and adapt. Given their conformational variability, one might guess that IDPs are poorly conserved at the sequence level, meaning that their amino acid sequences vary substantially between proteins or across species. To a large extent, this is true. Nevertheless, hidden within these seemingly divergent sequences are conserved short linear motifs (SliMs), which are recurring sequences only a few amino acids long that are essential for controlling particular protein functions. A wide variety of SliMs exist within IDRs, often at conserved positions. To name a few, the FG motif, consisting of just two amino acids, phenylalanine and glycine, is repeatedly found in IDRs of the protein nucleoporin, where these motifs contribute to the selective trafficking of proteins through the nuclear pore complex. Likewise, the KR clusters are clusters enriched in lysine and arginine and function as nuclear localisation signals, directing proteins towards the nucleus on demand (3). Interestingly, IDPs’ strategy parallels the broader and better-known process by which multicellular organisms become more complex, where functions become increasingly distributed among smaller, more specialised units, rather than being performed by the single, massive entity. One emerging view is also that these SliMs work cooperatively in a concerted way to create interactions, much like the division of labour in larger biological organisations (3). In addition to their flexibility, another major benefit of this increased complexity is evolutionary efficiency. In general, IDPs tend to evolve faster than ordered proteins. These short, abundant SliMs occur within flexible disordered regions and can easily emerge through just one or two point mutations from nowhere. Consequently, instead of having to evolve the entire complex 3D protein structure, evolution simply has to generate a short linear motif within IDR with point mutations - a kind of evolutionary shortcut to acquiring new functions and growing complexity (4). Controlled chaos At high concentrations, IDPs can quickly become harmful, engaging in misinteractions with inappropriate target proteins. It easily creates chaos through their powerful ability to form transient but strong interactions with multiple proteins simultaneously. How, then, can we rely on disorder for critical life functions? Why can evolution afford this amount of chaos, and is it worth it to trade certainty for greater complexity? The central answer lies in the word ‘control’. In general cases, cells exercise tight spatiotemporal control on IDPs, regulating them to ensure they exist in the right place, at the right concentration, and for the right duration. This regulation begins far before the protein is produced, starting with mRNA transcripts. Transcripts encoding IDPs exhibit higher decay rates, and are disproportionately targeted by small, non-coding RNAs that control translation. Once translated, IDPs are then subject to dynamic fine-tuning. Phosphorylation, for instance, is proposed to alter its net charge and binding affinity, modulating its stability and localisation on demand. On the other hand, IDPs undergo much more ubiquitination (attachment of an ubiquitin protein tag) compared to folded proteins, which renders them inherently sensitive to ubiquitin-proteasomal degradation pathways (5, 6). Evolution can therefore afford the apparent chaos of disorder because of the multiple checkpoints that tightly regulate the localisation, abundance and lifespan of IDPs. By doing so, evolution can trade for more efficient and complicated life processes more safely. The kaleidoscope of biological order created by IDPs The tension between disorder and order also appears across a remarkable kaleidoscope of IDP functions. Some IDPs help misfolded proteins in refolding by transiently binding exposed regions and compensating for the unfavourable free energy increase with stabilisation of the intermediates. Others act as entropic chains, maintaining the spatial organisation of protein domains, or as molecular assemblers, engaging multiple binding partners simultaneously to promote formation of higher-order protein complexes, including ribosomes. Still others act as molecular scavengers, storing and neutralising molecules or compounds harmful in large amounts, such as reactive oxygen species and heavy metals. Taken together, these functions suggest that biological order can often arise through disorder. Is biological order even possible without the presence of disorder then? The multitude of functions dependent on IDPs’ flexibility may have shown us that an ever-changing kaleidoscope of uncertainties and possibilities is sometimes what really underpins the appearance of tightly regulated biochemical processes. The D² concept - Disorder in disorder Disorder is, however, still disorder, and its flexibility makes it hard for cellular mechanisms to 100% guarantee that it is under control. One prominent risk of IDPs is their propensity to form amyloid fibres, structures notorious for their associations with neurodegenerative diseases such as Alzheimer’s, Parkinson’s and Huntington’s disease. While folded proteins have to undergo thermodynamically unfavourable unfolding or partial-unfolding before transitioning into amyloid states, IDPs already access aggregation-prone conformations. Underlying their delicate balance between flexibility and functionality is the capacity to self-assemble into β-sheet-rich structures, which can progressively aggregate into insoluble, elongated amyloid fibres through various mechanisms. As amyloid-like structures are highly thermodynamically favourable, once it has progressed, it is difficult for IDPs to get out of it (7, 8). The consequences of loss of control extend beyond amyloid fibril formation because of the ubiquity of IDPs in signalling and regulatory networks. The intricate nature of biochemical processes naturally makes a single fault significant. Dysregulation of IDPs can disrupt and redirect biochemical pathways. For instance, prolonged cellular stress can result in an IDP-mediated protective pathway being altered into one that triggers programmed cell death. Cancer cells can likewise exploit dysregulated IDP-mediated pathways to tolerate elevated levels of reactive oxygen species and facilitate metastasis (8). Evolution also doesn’t only benefit humans. By exploiting the flexibility of IDP, bacterial pathogens can mimic host proteins with surprising precision, hijacking the host’s own biological machinery for their benefit. In some cases, a bacterium needs little more than a tiny IDP with a few strategically positioned amino acids, to seize control of a host enzyme and rewrite its function. Like a puppeteer pulling the strings of a marionette, the pathogen redirects the enzyme’s activity to serve its own interests, helping it survive, thrive and persist within the host (9). Evolution neither guarantees flawless solutions nor strives towards a predetermined destination. Biological systems, too, have no fixed formula of how much chaos and control to introduce, only a ceaseless process of balancing chaos and control. Nevertheless, uncertainty never negates the need for change, and nature never stops bringing uncertainty to life. Chaos is brought to the world by evolution, just like courage is brought, like the first flame flickers to life on a barren earth and the first words stumble tentatively from the mouth. None of them are perfectly safe in itself. It is through balanced use that each finds its right place to be. References Uversky VN, Kulkarni P. Intrinsically disordered proteins: Chronology of a discovery. Biophysical Chemistry . 2021;279:106694. doi: 10.1016/j.bpc.2021.106694 Trivedi R, Nagarajaram HA. Intrinsically Disordered Proteins: An Overview. Int J Mol Sci . 2022;23(22):14050. doi: 10.3390/ijms232214050 Nguyen Ba AN, Yeh BJ, van Dyk D, et al. Proteome-Wide Discovery of Evolutionary Conserved Sequences in Disordered Regions. Science Signaling . 2012;5(215):rs1-rs1. doi: 10.1126/scisignal.2002515 Pajkos M, Dosztányi Z. Chapter Two - Functions of intrinsically disordered proteins through evolutionary lenses. In: Uversky VN, ed. Dancing Protein Clouds: Intrinsically Disordered Proteins in the Norm and Pathology, Part C . Vol 183. Progress in Molecular Biology and Translational Science. Academic Press; 2021:45-74. doi: 10.1016/bs.pmbts.2021.06.017 Babu MM, van der Lee R, de Groot NS, Gsponer J. Intrinsically disordered proteins: regulation and disease. Current Opinion in Structural Biology . 2011;21(3):432-440. doi: 10.1016/ j.sbi .2011.03.011 Edwards YJ, Lobley AE, Pentony MM, Jones DT. Insights into the regulation of intrinsically disordered proteins in the human proteome by analyzing sequence and gene expression data. Genome Biol . 2009;10(5):R50. doi: 10.1186/gb-2009-10-5-r50 Uversky VN, Oldfield CJ, Dunker AK. Intrinsically disordered proteins in human diseases: introducing the D2 concept. Annu Rev Biophys . 2008;37:215-246. doi: 10.1146/annurev.biophys.37.032807.125924 Dormann D, Lemke EA. Adding intrinsically disordered proteins to biological ageing clocks. Nat Cell Biol . 2024;26(6):851-858. doi: 10.1038/s41556-024-01423-w Forni D, Cagliani R, Sironi M. Unstructured but dangerous—the emerging roles of intrinsically disordered protein regions in bacterial pathogenesis. FEBS Letters . 2026;600(18):2695-2715. doi: 10.1002/1873-3468.70368 < Previous Back to Issue 11 Next >

< Back to Issue 11 Two Beginnings, One Body by Kara Miwa-Dale 27 October 2026 Illustrated by Michelle Van Edited by Florine Callista Note: This story is fictional but draws on real events; in particular, the case of Karen Keegan, described by Yu et al. (2002). It was an ordinary Tuesday when we got the results. “You are not the mother.” I laughed. Not because it was funny, but because the alternative was just too ridiculous to take seriously. “The test is wrong,” I said. “Do it again.” “We already did.” The geneticist sat opposite me, choosing her words carefully. “The results are consistent with your husband being the father,” she continued. “But we haven’t found the genetic markers we’d expect to see from you.” I stared at her, then looked at my husband. He’d been there for everything: the first ultrasound, the strange cravings, the swollen ankles, the sleepless nights. He’d held my hand through labour and watched our daughter enter the world, her first cry erupting from her tiny lungs and filling the room. I had carried her for nine months. I knew I was her mother. My first instinct was to look for a mistake. Had the samples been mixed up? Had someone put the wrong label on the tube? Could there have been contamination, or was there an error in the laboratory? Surely, there had to be an explanation. The geneticist explained that the laboratory would first review the sample identification, records and quality controls. The original findings needed to be checked before anyone could draw conclusions. Depending on the circumstances, new samples might be collected and tested. I clung to the possibility that something ‘ordinary’ had gone wrong. But the discrepancy remained. I wanted to understand what the test had actually found. The geneticist explained that a parentage test compared particular regions of DNA, looking at genetic markers that a child would ordinarily inherit from their biological mother and father. A child receives roughly half their nuclear DNA from each parent, and the markers examined should be consistent with that inheritance. My husband’s genetic profile was consistent with fatherhood. Mine wasn’t. If my daughter had inherited half her DNA from me, where was it? I’d always thought of DNA as like a fingerprint – something fixed and consistent throughout the body. A sample of blood should identify the same person as a swab from inside the cheek. I’d never had any reason to question that assumption. But the human body can be more complicated than we imagine. Chimerism is a rare condition in which a person carries two or more genetically distinct populations of cells. The name originates from the Greek mythology, Chimera — a creature made up of different animals. In humans, the reality is less monstrous but no less extraordinary: a single person can carry two or more genetically distinct populations of cells. I struggled to understand how someone could live an entire life without knowing. Was it possible that my own body held something like this? One form, called tetragametic chimerism , can occur when two eggs are fertilised by two different sperm. Ordinarily, each fertilised egg develops into a separate embryo, potentially becoming fraternal twins. Very rarely, the two early embryos fuse, and their cells combine to develop into a single person. The person may carry two distinct genetic lineages, each originating from a different fertilised egg. Four gametes, two beginnings, one body. Someone with chimerism may grow and develop normally, with no outward signs that anything is unusual. The two genetic lineages can be distributed differently throughout the body. A blood sample might predominantly reveal one, while other tissues contain evidence of another. The possibility was extraordinary. But the doctors needed evidence. They arranged for DNA from additional tissues to be examined. By comparing genetic profiles from different sources, they could investigate whether my blood told the same story as the rest of my body. It didn’t. The genetic profile found in my blood did not fully match those from other tissues. Further analysis was needed to establish whether the differences reflected two distinct genetic lineages rather than technical problems. Eventually, the tests confirmed it. I carried two genetically distinct cell populations, with profiles consistent with two separate fertilised eggs having contributed to my body. I had tetragametic chimerism. The words seemed almost impossible to reconcile with the life I’d lived. All my life, I’d assumed I had one genetic identity. Yet my body carried evidence of two beginnings, and I’d grown up without the slightest idea. At first, the discovery felt like an answer. It explained why the DNA in my blood had failed to show the expected genetic relationship with my daughter. But it also raised a question far more frightening than anything I’d considered before. What did this mean for my daughter? If my body contained two genetic lineages, which one had contributed to my daughter? The discovery of chimerism could explain why the original test had failed to identify the expected maternal relationship. But it could not, by itself, answer every question about the DNA I had passed on. Further investigation would be needed to determine whether my daughter had inherited genetic material from my second lineage. For the first time, though, the results offered an explanation for something that had seemed impossible. The test hadn’t necessarily been wrong. It had analysed the DNA in the sample it was given, but couldn’t capture the full complexity of my genetic makeup. I was still the woman who had carried my daughter, given birth to her and brought her home. None of that had changed. But now I understood something I had never imagined possible: the DNA we’d tested wasn’t the entire story. Bibliography Semikhodskii A, Makarova T, Sutyagina D. Paternity pseudo-exclusion caused by tetragametic chimerism in a gestational surrogacy case. European Journal of Medical Genetics. 2023 Aug 1;66(8):104799. https://doi.org/10.1016/j.ejmg.2023.104799 Yu N, Kruskall MS, Yunis JJ, Knoll JH, Uhl L, Alosco S, Ohashi M, Clavijo O, Husain Z, Yunis EJ, Yunis JJ. Disputed maternity leading to identification of tetragametic chimerism. New England Journal of Medicine. 2002 May 16;346(20):1545-52. https://doi.org/10.1056/NEJMoa013452 Yu Q, Li Q, Gao S, Su Y, Deng Z. Congenital tetragametic blood chimerism explains a case of questionable paternity. Journal of Forensic Sciences. 2011 Sep;56(5):1346-8. https://doi.org/10.1111/j.1556-4029.2011.01794.x < Previous Back to Issue 11 Next >

< Back to Issue 11 Neurotechnology: The Technology Decoding the Brain by Luca Vithani 27 October 2026 Illustrated by Yolanda Huang Edited by Aimee Fogarty-Bennett Krista and Tatiana Hogan are craniopagus conjoined twins, which means that they are physically connected at the skull. In this rare case, not only are their skulls conjoined, but so are their brains. The thalamus functions as the central information relay hub in the brain. You can think of it as the brain’s post office: processing and routing most motor and sensory information to different regions of the brain. The connection of each of the twins’ thalamus creates what is called a Thalamic Bridge. MRI imaging showed bidirectional neural transmission through this bridge. When they undertook hospital tests, flashing a light in one twin's eyes produced a response in the other's visual cortex, even when her own eyes were covered. Further reports say that Krista and Tatiana share taste, share touch, can see out of each other's eyes, and the twins say that they can have conversations with each other inside their own heads. This demonstration of neural integration raises questions about the potential of neurotechnology and makes you wonder how your mind and consciousness could be shared with another person or even technology (1,2). The human brain is considered the most complex object known to science. Neurotechnology has become one of the best ways to observe and understand the rapidly changing, complex patterns and signals of the brain. The field is accelerating faster than it ever has and is entering a period of explosive growth. At its core, neurotechnology is any technology designed to record, analyse, or modulate neural activity. This includes devices/equipment from cochlear implants, to brain-computer interfaces. A brain-computer interface (BCI) is a direct communication link between the brain's electrical activity and an external device, like a computer or robotic hand. You can simplify the majority of BCIs into three components: The probe: The part that directly connects to, retrieves signals from, or stimulates the brain. The signal processing and decoding layer: Turns neural activity into an interpretable signal. The software and application layer: This is where the decoded signals get applied, analysed, or used for feedback (visual, haptic or direct stimulation). A BCI company called Science Corp, founded by former co-founder and president of Neuralink, Max Hodak, is working on a very interesting new type of probe called a Biohybrid Probe. Most traditional BCIs plug into the brain with electrodes, which can cause initial and continuing damage to the surrounding neurons. This means that the signals the probes are initially trying to read become degraded, and degrade even more over time. Science Corp believes the solution may be a biohybrid probe. The idea is that the probe uses engineered lab-grown neurons which are embedded in electronics and then engrafted onto the brain. The goal is that these neurons naturally grow into the brain, creating a biological bridge that can then transmit and receive signals to and from an external device. Although it is still in its early phases, there have been promising results shown across a variety of studies. They have tested the idea on mice and saw the transplanted neurons sending out fibres down into the brain of the mice. Some of these neurons also showed electrical activity, similar to the normal brain cells in the mice, suggesting they were behaving like functioning neurons and integrating with the mice's brain (3). The founder of Science Corp likened the concept to having a “13th cranial nerve”; an additional neural pathway extending the brain's ability to connect with the external world. He also compared the idea to the movie Avatar, where the native people named the Na’vi have a ‘neural braid’ that allows them to form a direct biological connection with other living beings and with ‘Eywa’, the interconnected consciousness of their world (4). As well as the early stages of their Biohybrid Probe, the company also has an established and clinically validated product called PRIMA. PRIMA is a retinal BCI implant that, in an almost biblical manner, is giving sight to the blind. It is specifically designed for people with age-related macular degeneration (AMD), which affects over 200 million people worldwide. Put simply, the user would wear special glasses with a small camera that captures images. The processed images are projected onto the implant using pulsed near-infrared light. The implant then electrically stimulates certain remaining retinal cells. Those signals travel through the optic nerve to the brain, which can then interpret the signals as visual information, helping someone see shapes, objects, and letters. Although it only partially restores eyesight to a patient, PRIMA is a major display of the potential of BCIs. In a major international clinical trial of patients with an advanced, late-stage form of AMD, more than 80% of participants regained meaningful central vision. In July 2026, PRIMA was approved for commercial use in 30 European countries. It has now become the first commercially approved BCI for vision restoration (5, 6). The capabilities of BCIs in general are becoming more established, with trials, studies and companies seeing immense progress. Some of the established capabilities of BCIs are: Thought to speech - A patient can think of a sentence such as “Hello World”, and a BCI can decode it into text or computer-generated speech (7). Controlling a computer with thought - A user can browse the internet, type, play video games, send emails, and much more (8). Controlling a robotic arm or device - A user could think “I want my hand to move left”, and a computer can then convert the brain signals from that thought into a command for a robotic arm. A paralysed woman has started painting with a robotic arm thanks to Neuralink’s brain implant (9, 10). All these capabilities aren't just mere one-time results or vague claims from researchers and companies. They are established, repeated capabilities that are becoming industry norms. For example, a 2026 study with a participant with Amyotrophic lateral sclerosis (a rare, fatal neurodegenerative disease that destroys the nerve cells responsible for controlling voluntary muscle movement) and severe paralysis demonstrated an implanted BCI being used for thought-to-speech for two years. The participant used it for more than 3,800 hours, producing more than 1.9 million words, with an average communication rate of 56 words per minute. In formal testing, attempted speech was decoded at more than 99% word accuracy (11). The potential benefits of this field are extraordinary, such as treatments for conditions that affect billions of people, from mental health conditions to epilepsy, Alzheimer's and the majority of brain-related conditions where treatment is hindered due to our lack of knowledge and control of the brain. We are already seeing the field of neurotechnology achieve feats like restoring sight to the blind, providing treatments for Parkinson's (12), as well as relieving some of the most intense suffering that results from conditions such as full-body paralysis. A majority of suffering is an experience mediated by the brain. Through advanced neurotechnology, we could radically reduce or even abolish the experience of suffering. Many argue neurotech may even be essential to fight the problem with ‘AI alignment’ by enabling us to augment human intelligence to keep up with or compete with AI. Either way, the field is in its early stages, with the brain being observed and decoded in more detail than ever before. References 1. Cochrane T. A case of shared consciousness. Synthese. 2021;199(1-2):1019-1037. doi:10.1007/s11229-020-02753-6. 2. The twins who share a brain [documentary]. In: Body Shock. Channel 4; 2010 May 25. 60 min. Available from: https://wellcomecollection.org/works/qugqdhqy 3. Brown J, Zappitelli KM, Dawson PM, et al. Optogenetic stimulation of a cortical biohybrid implant guides goal directed behavior [preprint]. bioRxiv. 2024. doi:10.1101/2024.11.22.624907. 4. S3. The Age of Neural Engineering (ft. Science Corp.) [Internet]. YouTube; 2024 Nov 23 [cited 2026 Oct 19]. Available from: https://www.youtube.com/watch?v=zIJi6dXN6Zc 5. Mathieson K, Loudin J, Goetz G, et al. Photovoltaic retinal prosthesis with high pixel density. Nat Photonics. 2012;6(6):391-397. doi:10.1038/nphoton.2012.104. 6. Holz FG, Le Mer Y, Muqit MMK, et al. Subretinal photovoltaic implant to restore vision in geographic atrophy due to AMD. N Engl J Med. 2026;394(3):232-242. doi:10.1056/NEJMoa2501396. 7. Card NS, Wairagkar M, Iacobacci C, et al. An accurate and rapidly calibrating speech neuroprosthesis. N Engl J Med. 2024;391(7):609-18. doi:10.1056/NEJMoa2314132. 8. Card NS, Singer-Clark T, Peracha H, et al. Long-term independent use of an intracortical brain-computer interface for speech and cursor control. Nat Med. 2026. doi:10.1038/s41591-026-04414-6. Available from: https://www.nature.com/articles/s41591-026-04414-6 9. Xu Y, Ding C, Shu X, et al. Shared control of a robotic arm using non-invasive brain-computer interface and computer vision guidance. Robot Auton Syst. 2019;115:121-9. doi:10.1016/j.robot.2019.02.014. 10. Neuralink. Creating art with the mind [Internet]. LinkedIn; 2026. Available from: https://www.linkedin.com/posts/neuralink_creating-art-with-the-mind-activity-7461117355343982592-ZptQ 11. Card NS, Singer-Clark T, Peracha H, et al. Long-term independent use of an intracortical brain-computer interface for speech and cursor control. Nat Med. 2026. Available from: https://www.nature.com/articles/s41591-026-04414-6 12. National Institute for Health and Care Excellence. Deep brain stimulation for Parkinson's disease [Internet]. London: NICE. Available from: https://www.nice.org.uk/ 13. Fulton JT. 9.1-2 An MRI image of the brains of the conjoined BC twin sisters [Internet]. ResearchGate; 2023. Available from: https://www.researchgate.net/figure/2-An-MRI-image-of-the-brains-of-the-conjoined-BC-twin-sisters-The-thalamic-bridge-is_fig10_372721878 14. Maclean’s. Conjoined twins share each other’s senses [Internet]. Available from: https://macleans.ca/news/canada/conjoined-twins-share-each-others-senses/ 15. IEEE Spectrum. Biohybrid BCI adds more neurons to the brain [Internet]. Available from: https://spectrum.ieee.org/brain-computer-interface-2671151260 16. Lightstorm Entertainment. Avatar [film]. 20th Century Studios; 2009. 17. Science Corporation. PRIMA retinal implant [photograph]. In: CBS News. Tiny eye implant, special glasses help some legally blind patients read again [Internet]. 2025 Oct 20. Available from: https://www.cbsnews.com/news/eye-implant-special-glasses-help-some-legally-blind-patients-read-again/ 18. Neuralink. Speaking with the mind [Video on the Internet]. YouTube; 2026 Sep 18. Available from: https://www.youtube.com/@neuralink < Previous Back to Issue 11 Next >

< Back to Issue 11 Many Disciplines, One Continent by Iris Gitlin 27 October 2026 Illustrated by Catherine Barnsley Edited by Ingrid Sefton Many Disciplines, One Continent: The People Behind Antarctic Science Research. Like many of us who were drawn to science degrees, I have a deep curiosity about the natural world. I’m not just interested in learning about things from a lab or from a screen; I want to go out and see how science impacts the world. But as I study, I can’t help but feel as though there isn’t much space for that in my degree. Like so many degrees, mine is very lab based and it left me wondering — what does it really look like to go out in the field and who are the people that do go out there? Antarctica, despite being one of the least inhabited environments on Earth, hosts scientists year-round. It’s an environment brimming with secrets that countless researchers are working to uncover — in summer, the number of scientists in Antarctica can reach 10,000 (1). Antarctica impacts global weather, ecosystems, and is even a great place for space research! Understanding such a complex environment needs the contribution of voices from every part of science. If there’s anywhere that hosts a diverse group of field researchers, it’s Antarctica. In other words, it’s the perfect place for my investigation! I decided to speak to three different scientists who have done fieldwork in Antarctica to see what I could learn about diversity in fieldwork. Katharina Hochmuth is a marine geophysicist. She has been involved in Antarctic research since her Master’s days and studies Antarctic ice-sheet development. Meanwhile, Yuhao Dai is a paleo-oceanographer. He’s interested in the chemical and physical compositions of the ocean in the past and how that influences today’s climate. Then, there is Tobias Staal, a geophysicist who researches glacial dynamics. His work also encompasses information theory, a branch of mathematics. While their research is different, all three have spent time in Antarctica doing field research. They agree that Antarctica is a particularly unique place to do field research because of its extreme environment. To even begin your research, you first have to reach Antarctica, which takes about a week from Hobart. Staal says the extreme conditions “mean a great deal of waiting: waiting for weather, aircraft, cargo, access, or safe conditions.” This is an aspect that’s unique to working in Antarctica: while most fieldwork requires some patience, working in Antarctica really teaches you to accept that you don’t have control over everything. Staal continued by remarking that “Sometimes you can spend an entire season without collecting much useful data.” Unlike in many other field locations, different scientists often travel and live together. Hochmuth describes this as a unique “community aspect” to the work, where scientists from vastly different backgrounds coexist. She says, “Most people are trying to support each other, rather than… ‘Ooh, this is my little island, and this is my kingdom.’” I’ve learned that this sort of collaborative environment really shapes the way that research happens in Antarctica. Antarctica is such a complicated environment, and it takes an incredible diversity of people to understand it in its entirety. During my conversations, the scientists told me stories about some of the fascinating people they met and learnt from during their field trips. “We had people from many different fields. So from biologists to physicists or physical oceanographers and chemical geographers and meteorologists,” Dai explained to me. He mentioned that one of the most exciting experiments he saw was scientists putting sensors on the heads of seals. The most unexpected part was that these scientists weren’t there to study the seals, they were using the sensors to study seawater properties! The seals were able to reach places that ships and other sensors would never be able to, a method that wouldn’t be possible without the collaboration of marine geophysicists and chemists with ecologists. It really struck me how many of the projects they told me about would not have been possible without the existence of interdisciplinary research. Hochmuth described the value of interdisciplinary research, saying it caused her to “make the connection that everything is interacting.” She encourages scientists to collaborate with each other, as “you may be able to generate some new ideas.” For Dai, working with scientists from different fields means that “you start looking at the question you're asking in a different light.” Working with people from different fields creates more complete scientists and more robust research. It can even be valuable to collaborate with non-scientists. Hochmuth commented that some of the most interesting people she met weren’t even scientists but members of the crew – like the cooks, the sailors and the people working on the deck. As she noted, “It's a deliberate choice to have a job like that. So they're always interesting individuals.” There are so many people that are needed to make trips to remote places successful – no matter what your background is, chances are there’s a place for you. My conversations made it clear to me that good fieldwork can only exist with the contributions of cooks, plumbers, writers and, of course, scientists from all different walks of life. As Staal puts it, “The most important problems cannot be solved by one discipline alone. Predicting the future of the Antarctic Ice Sheet, for example, requires observations of ice, ocean, atmosphere, and geological history, alongside modelling and computation. Communicating those predictions and mapping out actions requires social science, humanities, law, and art.” The more I spoke to Hochmuth, Dai and Staal, the more I understood the value my own scientific background could have in field research. I became curious to learn how to become an Antarctic scientist myself (as I’m sure many of you are as well!). When asked what they would say to young scientists trying to follow in their footsteps, the main advice they had to give was to make connections and approach people with confidence – the opportunities are found through speaking to other people involved in this work. As Hochmuth puts it: “If you don't tell anyone you're interested in it, it will never happen.” Staal’s reflection on this question really stuck with me: “Do not make fieldwork the goal. The objective is not to find a way to go to Antarctica; the objective is to do good science, always and anywhere.” Good science is diverse and may lead you to all sorts of places, Antarctica or elsewhere in the world. However, Staal finished by saying that if your research does happen to take you to the field, “Learn how to operate a ratchet strap!” Some organisations suggested by our scientists that support early career Antarctic scientists: https://antarcticsciencefoundation.org/ https://scar.org/ https://www.antarctica.gov.au/ References The British Antarctic Survey. Antarctic Factsheet & Geographical Statistics [Internet]. 2026 [cited 2026 Oct 23]. Available from: https://www.bas.ac.uk/about/education-and-schools/antarctic-factsheet/ < Previous Back to Issue 11 Next >

< Back to Issue 11 Coral Holobionts: A Secret Society under the Waves by Anjika Shankar 27 October 2026 Illustrated by Esme MacGillivray Edited by Luci Ackland From early classifications marking them as stones, to the gradual understanding of their plant-like morphology, you could say that corals have had a colourful history. In fact, medieval scholars from the 1st century AD used to believe that corals were ambiguous entities caught between plants, animals and minerals (1), explaining a possible evolutionary connection between the three, until prominent natural scientist William Herschel finally revealed their animal cell composition in the 18th century (2). But if not hardened ‘underwater flowers’ or some sort of ‘boundary organism’, what even are corals? Essentially classified under the kingdom Animalia, corals are further categorised under the phylum Cnidaria (the same phylum to which jellyfish belong) and class Anthozoa, as they are considered aquatic invertebrate animals. They are made up of tiny animals called polyps , each with a mouth surrounded by tentacles and a central digestive cavity. Most reef-building corals secrete a hard skeleton of calcium carbonate (CaCO₃). Over generations, the accumulated skeletons of colonies form the foundation of coral reefs, creating habitats for an extraordinary diversity of marine life (3). So, when we’re talking about ‘coral reefs’…what exactly are we referring to? The hardened calcium-base structure? The fragile polyps inside? Neither. We’re referring to the countless partnerships that collectively sustain our Earth’s most biodiverse ecosystems: the coral holobiont . Each coral ‘colony’ can host a community of microorganisms, including bacteria, archaea, fungi, viruses and microscopic algae. Together, the coral and its associated organisms form a holobiont : an ecological partnership in which multiple forms of life coexist and interact (4). The combined genetic information of the host and its associated organisms is called the hologenome . Whether this collective should be considered a single unit of evolution remains debated, but the concept highlights an important point: coral health cannot always be understood by studying the animal alone (4). Symbiosis: Partnerships built on mutual benefits One of the best-known symbiotic relationships corals have is with photosynthetic dinoflagellates from the family Symbiodiniaceae . These microscopic algae live within coral tissues and use sunlight to convert carbon dioxide into organic compounds through photosynthesis. The coral receives some of this energy-rich material, helping support its metabolism, growth and skeleton formation. In return, the algae gain shelter, access to nutrients and a relatively stable environment (5). Many other members of the coral microbiome may contribute to nutrient cycling, chemical defence and the breakdown of organic matter. However, these functions differ among coral species and microbial communities; the mere presence of a microorganism or a particular gene doesn't automatically prove that it benefits its coral host (4). The ‘partnership’ between corals and microorganisms isn’t simply a fixed arrangement, but a network of exchanges shaped by the surrounding environment. Light, temperature, food availability and water chemistry can alter which organisms thrive and how they interact. Therefore, two corals that look similar may depend on very different biological relationships. Two corals, two ways of life Staghorn corals (Acropora millepora): Powered by sunlight Acropora millepora, otherwise commonly known as staghorn coral, is a branching, reef-building coral found in shallow tropical waters, mostly in the western and central Indo-Pacific. Its relationship with Symbiodiniaceae illustrates how photosynthesis can power life in nutrient-poor environments. Sunlight penetrates the shallow water, allowing the dinoflagellates to fix inorganic carbon into organic compounds like glycerol, glucose, and amino acids. Some of these compounds are transferred to the coral, providing an important source of energy, and supporting growth and calcification (5). Image 1: Acropora millepora photographed near the Una Una island of Indonesia (6) The coral, in turn, offers the algae a protected place to live, and access to carbon dioxide and nutrients produced through the host’s metabolism. This exchange helps explain how tropical reefs can flourish in waters where nutrients are often scarce. Yet, there exists a glaring vulnerability: when seawater becomes excessively warm, their symbiosis can be disrupted. The coral may lose or expel its algae, or their pigments, causing the colony to appear white in a process known as coral bleaching (5,7). Eye corals (Desmophyllum pertusum): Deep down in the dark Far below the sunlit surface, the cold-water coral Desmophyllum pertusum , commonly known as eye coral but formerly known as Lophelia pertusa , builds reefs in darkness. Unlike many tropical reef-building corals, it doesn't rely completely on photosynthetic Symbiodiniaceae . Instead, it obtains energy by capturing zooplankton and organic particles from the surrounding water (8). Image 2: Desmophyllum pertusum photographed near the Swindle island of Canada (9) For the eye coral, like other deep-sea corals, its microbiomes include bacteria from groups historically classified as Bacteroidetes, Firmicutes and Actinobacteria. Also, research comparing deep-water and tropical coral-associated microbial communities found a greater representation of genes associated with chitin degradation in the deep-water communities examined (10). Chitin forms part of crustacean exoskeletons, so these genes suggest the potential to break down material from marine organisms. However, it isn’t fully clear whether those particular strains from all three groups directly digest crustacean remains for the coral’s benefit or for their own. A further possibility emerges where cold-water corals occur near methane seeps , where microbial communities can use chemical energy rather than sunlight. Some sulfur-oxidising bacteria obtain energy by oxidising reduced sulfur compounds and use it to fix inorganic carbon into organic molecules. Research suggests that certain deep-sea corals may associate with these microbes, but the strength and significance of these relationships vary between species and locations (11). For D. pertusum , the contribution of seep-derived carbon to nutrition remains uncertain; thus, it shouldn’t be assumed that this coral routinely depends on sulfur-based chemosynthesis (11). Image 3: Schematic summary of the symbiotic associations among Symbiodiniaceae, bacteria, and hosts in warm-water coral (left) and cold-water coral (right) holobionts. Coloured lines represent different metabolic pathways of Symbiodiniaceae (green), bacteria (red), and hosts (black), and line thicknesses represent relative frequency of different metabolic pathways’ usage (12) As illustrated in the image above, together, these two corals reveal different ways of life: one sustained by a sunlight-powered symbiosis with photosynthetic microalgae, the other by filter food capture and bacterial processing in the deep, dark sea, with possible additional chemical-energy pathways in seep-like environments. In all of this, the holobionts also provide shelter, nutrition and other resources to macroorganisms like marine animals and plants. Then, what makes an individual? If a coral’s survival depends on organisms living within and around it, where does the individual begin and end? Is a coral the animal itself, or the broader community that sustains its survival and functionality? Holobionts as a concept encourage scientists to consider the coral as part of a changing biological network . The organisms within that ecological network aren’t necessarily permanent, nor do they always share identical interests. Microbial communities can change with environmental conditions, and some associations may be beneficial in one context but less useful in another (4). A coral is, therefore, both an organism in its own right, and a participant in a wider ecological system. This complexity makes the idea of biological individuality less like a hard boundary, and more like a shifting mosaic. How do holobionts get disrupted, and what happens then? Climate change is placing coral partnerships under increasing pressure. During marine heatwaves, elevated temperatures can disrupt the coral–algal relationship and trigger bleaching. Bleaching doesn't necessarily mean a coral is dead: if conditions improve quickly enough, surviving corals may regain their symbionts and recover in time. If stress is intense or prolonged, however, the entire reef may starve, become more vulnerable to disease, or die (5,7). A recent investigation of living and fossil corals from the Galápagos Islands shows how reefs can preserve evidence of past environmental shocks. By analysing chemical signatures in coral skeletons, researchers reconstructed patterns of El Niño and La Niña over a long period. Their findings indicate that El Niño events during the past four decades were stronger than those recorded in the sampled millennium, with the intensification occurring alongside global warming (13). This association is concerning because El Niño can bring unusually warm waters and weaken nutrient-rich upwelling in the region. The coral cores also contained “ death horizons ”: layers of calcified material produced by other reef organisms, including algae, gastropods and barnacles, that interrupted the coral skeleton during periods of mortality. Despite that, in some cases, coral growth later resumed over these surfaces (9). These layers reveal that reef recovery is possible, but not guaranteed. So, a surviving skeleton can provide a surface for new growth, but a reef’s future still largely depends on whether living corals can settle, grow and withstand further stress. Moreover, other pressures compound the problem. Sediment runoff can cloud water, reduce light and smother coral tissue. Changes in ocean circulation can alter temperature, food supply and nutrient availability. Deep-sea mining may damage fragile cold-water coral habitats directly or disturb sediments that affect nearby ecosystems. Together, all these stressors can alter not only coral survival, but also the microbial communities and ecological relationships on which corals depend. Understanding the holobiont and protecting it Conservation researchers are exploring assisted evolution, selective breeding and the use of heat-tolerant algal partners to improve coral resilience, because understanding coral holobionts helps us know the nature of the interactions among various species. Microbiome manipulation, including introducing potentially beneficial microbes, is another area of investigation, although its effectiveness and ecological safety require further testing (14). These approaches may contribute to restoration, but they cannot replace our collective effort in reducing greenhouse gas emissions and protecting reefs from local stressors. Furthermore, corals don’t need soil to establish a reef. Their planulae (coral’s larvae) can settle on suitable hard surfaces, including the calcium carbonate skeletons of earlier generations. As colonies grow, they build upon this foundation; other organisms, both micro and macro, can also colonise and reshape it. In a beautiful yet grim way, a coral’s skeleton is both a record of past life and a platform for future communities (3, 13). Unfortunately though, a dead reef structure alone doesn't guarantee recovery: successful regrowth depends on living corals, suitable conditions and the return of ecological relationships. All in all, coral holobionts reveal that life is rarely an isolated journey. From tropical, sunlit branches to cold, dark seafloors, corals survive through different combinations of inter- and intra-species relationships, resources and environmental conditions. Like a kaleidoscope, each reef forms a distinctive pattern from many interacting parts. Protecting these ecosystems means preserving not only the coral animals, but also the intricate relationships that allow reefs to grow, adapt and endure while providing a place to sustain life for other organisms of varying shapes and sizes. References Why have corals long baffled scientists? [Internet]. Fondation Tara Océan. 2026 [cited 2026 Oct 23]. Available from: https://fondationtaraocean.org/en/scientific-news/the-history-of-science-and-coral/ Herschel W. The scientific papers of sir William Herschel. 1912. National Oceanic and Atmospheric Administration. What are coral reefs made of? Silver Spring (MD): NOAA; [cited 2026 Oct 19]. Available from: https://oceanservice.noaa.gov/facts/coralmadeof.html van Oppen MJH, Blackall LL. Coral microbiome dynamics, functions and design in a changing world. Nat Rev Microbiol. 2019;17(9):557-567. doi:10.1038/s41579-019-0223-4. Roth MS. The engine of the reef: photobiology of the coral-algal symbiosis. Front Microbiol. 2014;5:422. doi:10.3389/fmicb.2014.00422. Acropora millepora acropora millepora or plate acropora [Internet]. Reeflex.net . 2026 [cited 2026 Oct 22]. Available from: https://reeflex.net/tiere/608_Acropora_millepora.htm Weis VM. Cellular mechanisms of Cnidarian bleaching: stress causes the collapse of symbiosis. J Exp Biol. 2008;211(Pt 19):3059-3066. doi:10.1242/jeb.009597. Roberts JM, Wheeler AJ, Freiwald A. Reefs of the deep: the biology and geology of cold-water coral ecosystems. Science. 2006;312(5773):543-547. doi:10.1126/science.1119861. dfopacific. Deepwater coral (desmophyllum pertusum) [Internet]. iNaturalist. 2019 [cited 2026 Oct 22]. Available from: https://www.inaturalist.org/observations/42866226 Pratte Z, et al. Functional gene composition and metabolic potential of deep-sea coral-associated microbial communities. Coral Reefs [Internet]. 2023 [cited 2026 Oct 19]. Available from: https://link.sprhttps://doi.org/10.1186/s40168-025-02254-zinger.com/article/10.1007/s00338-023-02409-0 Vohsen S, et al. Deep-sea corals near cold seeps associate with sulfur-oxidizing chemoautotrophs in the family Ca. Thioglobaceae. Microbiome [Internet]. 2025 [cited 2026 Oct 19]. Available from: https://link.springer.com/article/10.1186/s40168-025-02254-z Gong S, Liang J, Jin X, Xu L, Zhao M, Yu K. Unfolding the secrets of microbiome (Symbiodiniaceae and bacteria) in cold-water coral. Kormas KA, editor. Microbiology Spectrum. 2023 Oct 17;11(5). Cole J. Galápagos corals show El Niño events are more extreme now than in past 1,000 years—here’s how my team made the discovery [Internet]. The Conversation. 2026 Aug 27 [cited 2026 Oct 19]. Available from: https://theconversation.com/galapagos-corals-show-el-nino-events-are-more-extreme-now-than-in-past-1-000-years-heres-how-my-team-made-the-discovery-272318 van Oppen MJH, Oliver JK, Putnam HM, Gates RD. Building coral reef resilience through assisted evolution. Proc Natl Acad Sci U S A. 2015;112(8):2307-2313. doi:10.1073/pnas.1422301112. < Previous Back to Issue 11 Next >

< Back to Issue 11 A Recipe for Discovery: Why Science Needs Different Perspectives by Jasmin Hirsch 27 October 2026 Illustrated by Ruby Cooke Edited by Ingrid Sefton One cook reaches for the salt, just as another adds a handful of herbs. Someone turns up the heat, while someone else insists it should be turned down. Then, halfway through, another decides the recipe needs changing altogether. Everyone has their own idea of what the dish needs, and before long, nobody is quite sure whose recipe they are following. There is a reason we say that “too many cooks spoil the broth.” In the kitchen, too many different approaches can end in disaster. In science, however, this might be exactly what a problem needs. A biologist, an engineer, a chemist and a computer scientist can look at the same problem and see something completely different. And that may be where discovery begins. Different cooks, different questions Scientific disciplines do not simply provide different sets of knowledge. They also teach people to look at problems differently. Faced with the same system, a biologist might ask, How does it work? An engineer might wonder , Can we build, improve or redesign it? A computer scientist might ask, Can we model or predict what it will do? A chemist may focus on which molecules could be changed, while a physicist might look for the forces and rules governing a system. None of these questions is necessarily the “right” one. Each illuminates a different part of the one problem. Imagine, for example, that human cells behave differently in a plastic dish than they do inside the body. A biologist might investigate whether signals from neighbouring cells are missing. An engineer, however, might question the dish itself: could we build the cells an environment that better resembles the human body? A physicist might ask whether forces, pressure or fluid flow are missing. The value of interdisciplinary research, then, is not simply I know something you don't. Sometimes, it is, You noticed a question about my problem that I didn´t know to ask (1) . Hearing the problem differently Hearing offers a striking example of this. Normally, sound vibrations entering the ear move tiny sensory hair cells inside the cochlea. These cells help convert mechanical vibrations into electrical signals that then travel along the auditory nerve to the brain (2). In severe sensorineural hearing loss, these sensory cells can be badly damaged, disrupting the transduction of sound waves to the brain. Understanding what has gone wrong is one problem. Finding a way around it requires another mode of thinking. If the auditory nerve is still functional and can transmit information, could the damaged sensory cells be bypassed? Could sound instead be converted into electrical signals and delivered to the auditory system directly? That idea sits at the heart of the cochlear implant. A microphone picks up sound, which is processed and translated into patterns of electrical stimulation. An array of electrodes inside the cochlea then stimulates the auditory nerve, bypassing much of the function of the damaged hair cells (3). But making that possible requires more than understanding the biology of the ear. Neuroscience contributes knowledge of how the auditory system processes electrical information. Engineering and signal processing are needed to transform sound into useful stimulation patterns. Audiology and psychology help researchers understand what implant users actually perceive and which sounds matter for communication and quality of life. Restoring access to sound is therefore not simply an engineering problem or a biological one. Different disciplines define – and solve – different parts of the same problem. From disease to drug The same principle can be seen in the search for new medicines. Imagine that researchers discover a particular protein plays an important role in a disease. Knowing this is useful, but it doesn't automatically translate to an usable drug. Instead, it creates a whole series of new questions – and different disciplines approach each one uniquely. While biologists investigate the protein’s role in the disease, structural biologists examine its three-dimensional shape to identify potential binding sites for drugs. Chemists use these insights to design and modify molecules that might interfere with the protein’s function. But with far too many possible molecules to make and test them all, computer scientists bring another approach to the table. Computational methods can help predict molecular properties, model interactions and identify promising candidates for laboratory testing (4). But a prediction on a computer is not medicine. Drug candidates still have to be synthesised, tested against their biological target, examined in cells and other experimental systems, and eventually assessed for safety and efficacy. Results from those experiments can, in turn, inform the next round of computational predictions and chemical modifications. Each discipline contributes a different approximation of the solution. Together, they can gradually turn an understanding of disease into something much more tangible: a potential treatment. When too many cooks really do spoil the broth Of course, putting more scientists around the same table does not automatically produce better science. Different disciplines have their own jargon, methods and standards of evidence. Even their definition of success can differ (1). An engineer may primarily want to know whether a device works reliably, while a biologist may also want to understand precisely why it works and which molecular mechanisms are responsible. There are practical problems too. Interdisciplinary projects may require specialised equipment spread across different laboratories, additional funding, different types of data and more people to coordinate (1). In other words, a bigger research kitchen can be harder to run. Yet the same differences that make collaboration difficult can also make it valuable. Explaining an idea to someone outside your field forces you to make assumptions explicit. A seemingly simple question can expose a blind spot. And where one discipline reaches the limits of its methods, another may offer an entirely different route forward. Let them cook Perhaps too many cooks really can spoil the broth, especially when nobody agrees on what they are cooking. Interdisciplinary research works not because more people automatically produce better answers, but because different perspectives can reveal different parts of the same problem. Science, after all, is not made by following a single recipe. Progress can begin when someone brings an unfamiliar ingredient, questions the instructions or suggests making something entirely different. The research kitchen may be noisier, messier and harder to coordinate when biologists, engineers, chemists and computer scientists gather around the same problem. But it also gains new questions, new tools and new possibilities. So don´t send the cooks away. Let them cook. References 1. Vladova G, Haase J, Friesike S. Why, with whom, and how to conduct interdisciplinary research? A review from a researcher’s perspective. Science and Public Policy. 2025;52(2):165-80. 2. National Institute on Deafness and Other Communication Disorders. How Do We Hear? Bethesda, MD: National Institute on Deafness and Other Communication Disorders; 2022 [Available from: https://www.nidcd.nih.gov/health/how-do-we-hear . 3. National Institute on Deafness and Other Communication Disorders. Cochlear Implants Bethesda, MD: National Institute on Deafness and Other Communication Disorders; 2024 [Available from: https://www.nidcd.nih.gov/health/cochlear-implants . 4. Sadybekov AV, Katritch V. Computational approaches streamlining drug discovery. Nature. 2023;616(7958):673-85. < Previous Back to Issue 11 Next >

< Back to Issue 11 Through the Looking-Glass by Edmond Sim 27 October 2026 Illustrated by Eric Wang Edited by Andrew Irvin "In another moment Alice was through the glass, and had jumped lightly down into the Looking-glass room." (1) Most of us, as children, have pressed a hand against a mirror and met another hand pressing back, one that is ours but not quite ours, with its left on our right. Consider how you know your own face. You have never seen it directly. You know it from mirrors, which reverse it, and from photographs, which flatten it. Remove the glass and you do not see yourself more truly. You do not see yourself at all. The world is no different. It reaches us through eyes, nerves, and a mind that shapes it before we see it. If we strip the shaping away, we would not see the world more clearly. Rather - we would not see at all. Alice’s experience in the Looking-Glass world, with its inverted reality and reversed logic, show that there exists a reality outside our perceived reflection. So the question we must ask is how much of what we see belongs to the world, and how much belongs to the glass. Philosophers have argued over it for millenia, and in the last hundred years physics has opened up this question once again. A Brewtiful Form David Brewster, a Scottish physicist working just after the turn of the 19th century, was investigating how light waves could be restricted to one plane (polarisation) by reflecting light back and forth between inclined reflective plates. In this process, while observing a candle, he noted down the “multiplication of the sectors formed by the extremities of the plates of glass” (2). Impressed by the beauty from the symmetry of the image – presumably a distraction from dark and miserable 19th century Scotland – he had a candle moment and his eyes lit up with pound sterling signs. Brewster saw the commercial side quickly. Seeing its potential in "all the ornamental arts" for its capacity to produce "an infinite variety of beautiful figures and patterns" (2), he took the setup, fixed it in a tube with pieces of coloured glass at the end and proudly displayed it at the Royal Society of Edinburgh. He built the name from three Greek words: καλός kalos (beautiful), εἶδος eidos (form) and σκοπέω skopeō (to look at) (2). A kaleidoscope was thus designed to be an instrument for looking at beautiful forms. Unfortunately, the patent did him little good. Before the licensed maker had stock ready, one of the instruments was shown to London opticians to take orders, and the opticians copied it. An estimated two hundred thousand kaleidoscopes sold in London and Paris in three months, and Brewster reckoned that at most a thousand of them had made them with his patent specification (2). Alas, he should’ve consulted an intellectual property lawyer. Its design is simple. At the far end of the tube sits the object cell, where loose fragments of coloured glass are held between a clear pane and a translucent one that diffuses the incoming light. Mirrors run the length of the tube, meeting along one edge. At the near end is the aperture. For the reflected copies of the cell to close up without a seam, the mirrors must meet at 180°/ n for some whole number n . At 45°, the eye sees eight sectors, and at 30°, it sees twelve (2). Brewster tried (and wrote about!) many possible configurations of his product. His patent favoured angles of 18°, 20° and 22½° (3), while later treatises specified 45°, 36° and 30° (2). Consider what the viewer does and does not see. While using the device, the glass fragments are never seen in the location where they actually lie; rather only their multiplied reflections reach the eye. The mirrors are never seen at all. They appear nowhere in the image, yet they fix the perceived shape and pattern. An argument I propose is that human perception operates like Brewster’s kaleidoscope: what we observe is not a raw, unmediated external world, but a structured projection produced by the interaction between external content (the inflow of ambient light) and internal architecture (glass beads and mirrors). Extending this analogy, the loose, coloured fragments of glass at the far end of the cylinder represent mind-independent matter. We never observe their unreflected physical form directly; however we observe their effects. The mirrors and lenses that produce the symmetrical patterns can be thought of as the inner workings of our neurology that permit us to process and perceive external stimuli. While Brewster's kaleidoscope postdates both Locke and Kant, I believe my proposed framework fits their theories of human perception well. Locke is concerned with the fragments: which qualities belong to objects themselves, and which exist only in the perceiver. Kant is concerned with the mirrors: what the mind contributes to every experience before any object appears in it. Neither theory was built solely from deductive philosophy alone, as each was modelled on the physics of its time. Locke's primary qualities (extension, shape, motion, number) are variables of a mechanics problem, drawn from the corpuscular philosophy of Boyle and Newton. Kant's forms of intuition are Newton's space and time, Euclidean, absolute and independent of each other, except drawn from the world into the mind. Finally, the relationship between physics and philosophy becomes strained when one of them is proven incorrect. We learnt that physical space could be curved, space and time merged into one, and particles no longer had a definite position and momentum at once. If Locke and Kant built on classical physics, to what extent can the framework of Brewster’s kaleidoscope persist today? Separating the chaff from the grain Locke's Essay Concerning Human Understanding (1690) contains one of philosophy's more pragmatic thought experiments: Take a grain of wheat, divide it into two parts; each part has still solidity, extension, figure, and mobility: divide it again, and it retains still the same qualities; and so divide it on, till the parts become insensible ; they must retain still each of them all those qualities (4, emphasis added). Locke’s mill is an instrument of division. Grinding changes a grain's size and shape, but the main point is that the mill cannot take away the fact that every piece has some size and some shape. Properties that survive every division must belong to matter itself. Locke's list of descriptors of this nature is “solidity, extension, figure, motion and number”. He calls them primary qualities, that are "utterly inseparable from the body, in what state soever it be" (4, emphasis added). Colour, taste, smell, warmth, and sound are not of the same vein. Locke uses the example of granite. If you turn off the light, its natural reddish-brown colour can’t be examined, and hence does not exist (4)! For these descriptors, he labels them as secondary qualities. Locke did not arrive at his list by introspection, rather deriving his theory of knowledge from the physics of the day. In the preface to his Essay he describes natural philosophers (a term that we would now call a physicist) as "master-builders," where he name-drops Robert Boyle and "the incomparable Mr. Newton," thus concluding that it is "ambition enough to be employed as an under-labourer in clearing the ground a little, and removing some of the rubbish that lies in the way to knowledge" (4). On this account, one can see that it is the philosopher's job to provide credence to the physicists’ theories. The physics that Locke was “clearing the ground” for was Boyle's. Boyle's Origin of Forms and Qualities (1666) set out the corpuscular philosophy, that all matter consists of tiny solid bodies, each with a size, a shape, and a state of motion, whereby every other property of an object comes from how its corpuscles are arranged within the body (5). This was not a new philosophy, but rather the rehashing of a truly ancient belief. In a rather terse remark that perhaps could have found its way into The Flintstones, Democritus put it more bluntly in the fifth century BC; "By convention sweet, by convention bitter, by convention hot, by convention cold, by convention colour; but in reality atoms and void" (6). However, what the seventeenth century added was the formalisation of mathematics. Galileo wrote that the book of nature "is written in the language of mathematics, and its characters are triangles, circles, and other geometric figures" (7). Descartes gave geometry an algebraic form, and Newton's Principia (1687) could predict where a body would be on the Day of Judgement from its mass, position and velocity (8). F = ma does not predict a body's smell. This is why Locke's list of primary qualities looks the way that it does. Size, shape, number and motion are the quantities that can be written into equations. Primary qualities are the variables of seventeenth-century mechanics. Galileo drew a hard conclusion about Locke’s secondary qualities. Tastes, odours and colours, he wrote, "reside only in the consciousness", so that "if the living creature were removed, all these qualities would be wiped away and annihilated" (7). Locke’s argument, however, was a little more nuanced. Rather than being a wishy-washy subjective experience unique to the observer, Locke writes that secondary qualities are " powers to produce various sensations in us by their primary qualities" (4, emphasis added). The redness of a rose is a real feature of itself; an arrangement of its particles that, given light and a human eye, produces the sensation of red. The sensation needs an observer. Its power does not. A power of this kind is a relation. It is defined not by what the rose is on its own, but by what it does to light and to eyes. Locke ranked such relational properties second, below the intrinsic properties a body has in isolation. Returning back to the kaleidoscope, the fragments at the end of the tube have a definite number and definite edges whether anyone is at the aperture. Those are its primary qualities. The colours that wash across our view are secondary; what the fragments' structure does to the light passing through them, and the effect that it has on the eye at the other end. The mirrors of the mind Kant thought Locke had stopped halfway. If colour belongs to how a thing appears rather than to the thing itself, why should shape and size be any different? In the Prolegomena (1783) he describes his completion of Locke’s theory: If I go farther, and for weighty reasons rank as mere appearances the remaining qualities of bodies also, which are called primary, such as extension, place, and in general space … no one in the least can adduce the reason of its being inadmissible (9). Kant’s main thesis is rather straightforward – everything we know about the fragments reaches us through the mirrors. We have never seen a fragment without them, so we have no way to verify which features belong to the fragment and which the reflection put there. Kant calls the thing as it is in itself ( ding an sich ), beyond any possible human experience as noumenon , and the thing as it appears to us as phenomenon . Locke's primary and secondary qualities are both forms of phenomena. Since the fragments in themselves cannot be known, Kant went to the piece of the toy that we may study from the inside: the mirrors. Kant claims that space and time are not features of the world we observe. Rather they exist as forms of our inherent intuition, the fixed structure through which anything must pass in order to be experienced at all. The existence and immutability of space and time are a priori ; not learned from empirical evidence, but rather an inherent logical deduction given to us at birth. We never perceive space itself, just as the viewer never sees the mirrors. Yet every experience we have is due to space. This is true for time. Kant’s illustrious success as a philosopher often overshadows his independent research in theoretical physics. His Universal Natural History (1755) used Newton's gravitation to explain how solar systems form (11). Similarly, we see Kant’s “plagiarism” of Newton’s theory of space and time, albeit placing his theory in a newfound context. In Newton’s Principia , "absolute space, in its own nature, without relation to anything external, remains always similar and immovable", and "absolute, true, and mathematical time, of itself, and from its own nature, flows equably without relation to anything external" (8). To the human mind, space and time are uniform and universal, and are able to be intuited from our human instinct. Consequently, space must be Euclidean, as our spatial intuition dictates "that the straight line between two points is the shortest" (10). Similarly, there must be a single time shared by every observer: "different times are only parts of one and the same time" (10). The question one must ask is: is our intuition true? Melted glass, bent mirrors Locke stopped the mill when the grain became “insensible,” and deduced the primary qualities from there. However, physics kept on grinding. Consider the following listed primary qualities: Bulk: Newton defined mass as a quantity of matter, something each corpuscle carries as its own. Upon grinding the wheat to its amylose and amylopectin starch chains, the molecules to atoms, the atoms to nuclei, and the nuclei to its protons and neutrons, we see that proton contains three quarks, but the quarks' own masses account for about one per cent of the proton's mass (12). The mass deficit is the energy of the strong nuclear force that binds the protons and neutrons together (13). Most of a proton's bulk belongs to none of its parts, rather the mass stems from relational energy from particles being in close proximity. Extension and figure: The electron has no measured size or shape. Experiments have proved an upper bound of 10 -18 metres, and the Standard Model treats it as a point particle (14). The "shape" of an atom is a variable probability distribution, with no fixed boundary that divides inside and outside. Motion : In quantum mechanics, a particle cannot have an exact position and an exact momentum at the same time, and consequently motion cannot be found out without losing where it is or where it is going (15). Some quantum interpretations, such as Bohm's pilot wave theory, have tried to restore definite positions by having a guiding wave “push” the actual configuration of particles, however this causes each particle's motion to depend on the positions of all the others (16). Locke's self-contained corpuscle does not agree with this. Solidity: Particles may experience quantum tunnelling effects, as the wave function representation of its position may result in a probability of finding the particle in a classically forbidden region. In practice, this is impossible due to the exponentially decaying probability of events happening for one, albeit an Avogadro’s number of particles. However, this still disproves Locke’s theory on the solidity of corpuscles. Number: Surprisingly, not even the comfort of invariant counting is spared. In quantum field theory, particles are excitations of underlying fields, and how many excitations are present can depend on the reference frame of the observer. An observer accelerating through what an inertial observer calls empty space may detect many particles (17). Every primary quality on Locke's list, probed to a microscopic scale, stops being an intrinsic property of a self-contained body and becomes a fact about relations; between quarks, between electrons, between a system and its measurement device. The fragments at the end of the tube, therefore, do not have a definite number or a fixed shape whether anyone is at the aperture. The mirrors are not spared from the physics revolution. In 1905, Einstein showed simultaneity is not universal; two observers in relative motion can disagree about whether two distant events happened at the same time, and neither is incorrect (18). In 1915, his general theory went further. The geometry of space and time is not fixed in advance but shaped by the matter and energy within it, and near mass geometry is no longer Euclidean (19). In Kant's kaleidoscope, the mirrors were fixed at one angle in every tube. Relativity dictates that the angle depends on how fast the tube is moving and the direction it is held. To make matters worse, the angle is not set by the viewer at all. The presence of external mass and energy determine the geometry, so the contents of the universe distort and bend the mirrors depending on where it is used. A Kantian can reply that ordinary experience still looks Euclidean, which indeed is true; as deviations locally are far too small to see. However, recall that Kant did not claim that space looks Euclidean. He claimed we know, a priori , and therefore with certainty, that if our intuition dictates it to be so, it is so. Mathematicians (Lorentz, Poincaré and the like) had already shown in the nineteenth century that non-Euclidean geometries are perfectly consistent and functional (20). Einstein showed that solutions to his field equations demonstrate that the space we occupy is non-Euclidean (19). Clearly, there is absolutely no hope for our kaleidoscope. Its fragments have melted and its mirrors have been bent, so our framework is futile, an antique relic that should best be left in the past? Rebuilding the kaleidoscope Let us first consider the glass fragments first. The history of physics is a graveyard of discarded things: caloric, phlogiston, the luminiferous aether. If our best theories have repeatedly been wrong about what exists, then who in their right mind would trust our newfangled current ones? Worrall answers that despite the incorrect precise detail of scientific theory, we still have the same mathematical structure persisting across time (21). He uses the example of light, where in 1820s Fresnel described light as a vibration in an elastic aether and derived equations for how much light a glass surface reflects and transmits for each polarisation. Maxwell's theory later replaced the aether with electric and magnetic fields. Whilst Fresnel was incorrect about what medium light vibrated in, his equations carried over unchanged. Now looking back at Locke's qualities, we see that he ranked intrinsic properties of corpuscles over relational powers. Modern physics found that at its core, particles do not possess intrinsic bulk and figure, but relations; group symmetries, conservation laws, the ways systems act on one another. Since Eugene Wigner's work in 1939, physicists have classified elementary particles not by size or shape but rather how they behave under the symmetries of spacetime, which is what fixes their mass and spin (23). Others argue that fundamental properties are powers in Locke's sense: that there is nothing more inherent to an electron's charge than rather what its charge allows it to do (24). So, a neo-Lockean keeps his method, but rather inverts Locke’s original order. Secondary, then primary. Now for the mirrors. Hans Reichenbach, writing in 1920 when general relativity was merely five years old, argued that Kant had combined two meanings of a priori (25). Its first meaning is that an a priori deduction is "necessarily true" or "true for all times." Evidently, an a priori belief that space is Euclidean has been proven necessarily false by Einstein’s relativity. However, its secondary meaning, and the one that Reichenbach claims can save Kant’s theory is "constituting the concept of object" (25). An a priori deduction must come after a framework before any test of its validity can mean anything. Michael Friedman develops this into what he calls the relativised a priori (26). Newton's mechanics presupposed Euclidean geometry and his calculus; Einstein's presupposed Riemannian geometry and the equivalence principle (gravitational mass equals inertial mass). Neither framework was learned from the experiments it made possible, since without it those experiments would have meant nothing. Yet each could be replaced when a scientific revolution replaced the whole system. Inside the tube So how much of what we see belongs to the world, and how much to the glass? There is no fixed share, because there is no barrier between them. Every observation is a physical interaction. Demanding to know what reality looks like without an observer is a logical fallacy; if you see a kaleidoscope from the outside without peering in, you wouldn’t know the pattern that emerges. What we can and should trust is invariance , the structural patterns that never change, no matter which angle you look from. Physics looks past the shifting reference frames of observers to find the conservation laws, mathematical symmetries, and field equations that hold firm across all of them. Yet we only uncover these invariants because we can each observe reality from our own kaleidoscope, sharing, testing, and validating what we see to map the relationships across the observable universe, both within and without. Nagel called the old ideal of objectivity "the view from nowhere" (27). Physics has swapped it for something attainable: the view that holds from everywhere. References Carroll L. Through the looking-glass, and what Alice found there. London: Macmillan; 1871. Brewster D. A Treatise on the Kaleidoscope . Archibald Constable & Co; 1819. Accessed September 27, 2026. https://archive.org/details/b29295440 Brewster D, inventor. A new optical instrument called "The Kaleidoscope" for exhibiting and creating beautiful forms and patterns of great use in all the ornamental arts. British patent 4136. July 10, 1817. Locke J. An Essay Concerning Humane Understanding . Vol 1. Project Gutenberg; 2004. Accessed September 27, 2026. https://www.gutenberg.org/ebooks/10615 Boyle R. The Origine of Formes and Qualities (According to the Corpuscular Philosophy) . Ric. Davis; 1666. Taylor CCW. The Atomists: Leucippus and Democritus. Fragments . University of Toronto Press; 1999. Galilei G. Discoveries and Opinions of Galileo . Drake S, trans. Doubleday; 1957. Newton I. Sir Isaac Newton's Mathematical Principles of Natural Philosophy and His System of the World . Motte A, trans; Cajori F, rev. University of California Press; 1934. Kant I. Prolegomena to Any Future Metaphysics . Carus P, trans. Open Court; 1902. Kant I. Critique of Pure Reason . Guyer P, Wood AW, trans. Cambridge University Press; 1998. Kant I. Allgemeine Naturgeschichte und Theorie des Himmels . Johann Friederich Petersen; 1755. Navas S, Amsler C, Gutsche T, et al; Particle Data Group. Review of particle physics. Phys Rev D . 2024;110(3):030001. doi:10.1103/PhysRevD.110.030001 Dürr S, Fodor Z, Frison J, et al. Ab initio determination of light hadron masses. Science . 2008;322(5905):1224-1227. doi:10.1126/science.1163233 Gabrielse G, Hanneke D. Precision pins down the electron's magnetism. CERN Courier . October 4, 2006. Accessed September 27, 2026. https://cerncourier.com/a/precision-pins-down-the-electrons-magnetism/ Heisenberg W. Über den anschaulichen Inhalt der quantentheoretischen Kinematik und Mechanik. Z Phys . 1927;43(3-4):172-198. doi:10.1007/BF01397280 Bohm D. A suggested interpretation of the quantum theory in terms of "hidden" variables. I. Phys Rev . 1952;85(2):166-179. doi:10.1103/PhysRev.85.166 Unruh WG. Notes on black-hole evaporation. Phys Rev D . 1976;14(4):870-892. doi:10.1103/PhysRevD.14.870 Einstein A. Zur Elektrodynamik bewegter Körper. Ann Phys . 1905;322(10):891-921. doi:10.1002/andp.19053221004 Einstein A. Die Feldgleichungen der Gravitation. Sitzungsberichte der Königlich Preussischen Akademie der Wissenschaften . 1915:844-847. Beltrami E. Saggio di interpretazione della geometria non-euclidea. Giornale di Matematiche . 1868;6:284-312. Worrall J. Structural realism: the best of both worlds? Dialectica . 1989;43(1-2):99-124. doi:10.1111/j.1746-8361.1989.tb00933.x Poincaré H. The Foundations of Science: Science and Hypothesis, The Value of Science, Science and Method . Halsted GB, trans. The Science Press; 1913. Wigner E. On unitary representations of the inhomogeneous Lorentz group. Ann Math . 1939;40(1):149-204. doi:10.2307/1968551 Bird A. Nature's Metaphysics: Laws and Properties . Clarendon Press; 2007. Reichenbach H. The Theory of Relativity and A Priori Knowledge . Reichenbach M, trans. University of California Press; 1965. Friedman M. Dynamics of Reason . CSLI Publications; 2001. Nagel T. The view from nowhere. New York: Oxford University Press; 1986. < Previous Back to Issue 11 Next >