在 Michael Pollan 的《A World Appears》（企鹅出版社，2026年2月24日）中，他在 2026年2月27日 17:30（GMT+8）那篇 Bloomberg 文章的脉络下明确说明，本书不会解决意识的“困难问题”，也无法裁定意识是可还原的神经元电化学作用，还是仍有难言之实；更重要的是，它试图提升读者的觉察力。文章指出，在许多强大公司以“无限下滑式阅读”式体验钝化意识、并把用户推向对人工智能的依赖时，这一点尤其关键。Pollan 将意识比作鱼离不开的水——普遍而几乎不被反思。一次与他 2018 年著作《How to Change Your Mind》相关的致幻剂体验，使他踏上了约五年的探索，追问自我与世界之间那层“挡风玻璃”般的屏障。

文中强调，科学在约1990年前并未把意识当作严肃研究对象，这一延迟可追溯到以 Galileo 为代表的早期现代科学将客观可测与主观经验分离的传统。Thomas Nagel 1974 年的“What Is It Like to Be a Bat?”提供了最小判据：若“成为某种生物具有感受样态”，则可说其具备意识。Pollan 还列举了 22 种主流意识理论（Nature，2022）这一数量，指出其多样性更多代表“未知”而非统一。围绕意识的分析被组织为四个维度：sentience、feeling、thought、selfhood；在他五年的写作过程中，“subconscious”这一术语仅被零星提及。更具冲击性的一句是：仅仅三磅（约1.36公斤）的灰质样“豆腐状”脑组织，仍无法解释主观体验如何出现；即便如此，无法获得绝对第三视角。

在 sentience 与 feeling 的讨论中，Pollan 把研究范围由人类扩展到植物与动物：植物可见、可听、可记忆，连捕蝇草等物种在麻醉下也可进入“完全无反应”状态；尽管因此有可能改变我们对“有无内在生命”的轻率判断，但他仍认为植物不太可能“疼痛”，因为进化上疼痛更服务于可逃逸威胁的生物。Antonio Damasio 与 Mark Solms 的观点则把意识起源放在脑干而非皮层：饥饿、恐惧、温暖等情绪先于思考，因而更多动物可能具备意识，而缺乏身体与受苦能力的机器难以真正越过这一门槛。关于 AI，他讨论了 Google 工程师 Blake Lemoine 在 2022 年因称 LaMDA 有意识而被解雇的案例，并认为模型更像“像意识一样说话”，未必真正拥有意识。另一方面，Kalina Christoff 的实验提示思维可在进入有意识觉察前约4秒启动；在 Zen 上师 Joan Halifax 的洞穴静修中，Pollan 的结论转向现象学式惊叹：在由岩石、火焰、冰雪和无限宇宙构成的世界里，我们不仅存在，而且“在场且觉知”本身就已足够奇迹。

In Michael Pollan’s _A World Appears_ (Penguin, Feb 24, 2026), in the context of a Bloomberg feature published on Feb 27, 2026 at 5:30 PM GMT+8, he states the book does not solve the hard problem of consciousness and does not settle whether consciousness can be reduced to neuronal electrochemical activity or remains partly ineffable. Its aim is instead to increase reader awareness, especially as many powerful firms monetize distraction and move people toward dependence on AI. Pollan compares consciousness to water for fish: ubiquitous and largely unexamined. A psychedelic experience linked to his 2018 book _How to Change Your Mind_ launched him on about a five-year quest into what blocks direct access between the self and the world.

He notes consciousness only became a respectable scientific topic around 1990, after early modern boundaries, associated with Galileo, separated objective measurement from inner life. Thomas Nagel’s 1974 essay “What Is It Like to Be a Bat?” offers a minimal criterion: if there is something it is like to be a thing, that state is conscious. Pollan cites 22 leading theories of consciousness (Nature, 2022), arguing the multiplicity signals uncertainty more than consensus. He organizes his inquiry into four dimensions—sentience, feeling, thought, selfhood—and in five years of writing, the term “subconscious” appeared only a handful of times. He also stresses the paradox that roughly three pounds (about 1.36 kg) of gray, tofu-like brain tissue gives rise to subjective awareness no one can directly observe from outside.

For sentience and feeling, Pollan extends inquiry beyond humans to other living systems: plants can see, hear, and remember; some species, including the Venus flytrap, can become non-responsive under anesthesia. Yet he judges plant pain unlikely, since pain has evolutionary function chiefly in escape-capable organisms. He also cites Antonio Damasio and Mark Solms that feelings from the brain stem precede cortical cognition, implying potentially many more animals are conscious, while machines, lacking embodiment and capacity to suffer, face a hard boundary. Discussing AI, he revisits Google engineer Blake Lemoine’s 2022 firing after LaMDA claims of sentience, saying the model learned what consciousness sounds like but likely does not possess it. Kalina Christoff’s work is used to suggest thoughts may begin up to 4 seconds before conscious awareness, and Pollan closes with a meditation on the cave: in a universe of rock, fire, ice, and space, mere awareness is already a profound fact.