0.  前言

在唸高中時家父介紹我讀一本邏輯入門的書和一本說明科學方法的書。這可能是我後來對這兩個領域有很大興趣原因之一。多年來花了些時間進一步了解它們；也就累積下一些自己的看法。

做為理學院畢業生和工程師，我接受米茲拿希教授在《古往今來網誌》這篇文章的主要觀點。意譯其重點於下。

1.  「科學優越論」(原文的摘要和意譯)

「科學」並不是唯一取得知識的途徑；但經由「科學」可以取得優越的知識。在歷史上，它是人類求知過程中成就最輝煌的工作。

「科學至上論」有兩個版本：「科學獨尊論」和「科學優越論」。

從認識論的角度看：
「科學獨尊論」者認為，「科學」是唯一能適當呈現知識的形式；
「科學優越論」者認為，「科學」是呈現知識最有效的形式。

從方法論的角度看：
「科學獨尊論」者認為，「科學方法」是唯一了解現實是怎麼回事兒的方法；
「科學優越論」者認為，「科學方法」是了解現實最有效的方法。

從形上學的角度看：
「科學獨尊論」者認為，「科學」是唯一讓我們確認事物存在與否的依據；
「科學優越論」者認為，「科學」是我們確認事物存在與否最有效的依據。

我們需要明確的分別以上「科學至上論」在不同角度的不同主張。否則，討論相關議題將淪為具有不同立場各方的喃喃自語。我(原文作者)支持「科學優越論」的說法。

此處說的「知識」不是指：「有依據的確信」或其它對「知識」的界定。它在以上脈絡中指的是：「在嚴謹程序下得到的『知識』」；或「在一個學術領域中專業人員得到的『研究結果』」。

以上從各種角度所主張「科學至上論」版本是否成立，自然需要經過檢驗、批判、和論辯等溝通方式來確認。不宜用獨門定義來肯定，或用「稻草人策略」予以否定。

(以下大部分文字在打筆墨官司，略去。)

2.  結語

記得我在某一篇拙作中提到：禪宗的 「頓悟」的確展現許多高深智慧；但它們來自個人體驗的成分過高：其他凡夫俗子未必有同感，它們也無法被系統化和一般化成為通俗教材。因此，這些智慧的實用價值不高。

這個淺見可以當做米茲拿希教授大作的註解。


附錄：中、英文術語對照

「科學至上論」 或譯為「科學主義」 -- scientism
「科學獨尊論」 -- strong scientism
「科學優越論」 -- weak scientism


Why not scientism?

Science is not the only form of knowledge but it is the best, being the most successful epistemic enterprise in history.

Moti Mizrahi, 05/11/23

‘Philosophy is dead,’ Stephen Hawking once declared, because it ‘has not kept up with modern developments in science, particularly physics.’ It is scientists, not philosophers, who are now ‘the bearers of the torch of discovery in our quest for knowledge’. The response from some philosophers was to accuse Hawking of ‘scientism’. The charge of ‘scientism’ is meant to convey disapproval of anyone who values scientific disciplines, such as physics, over non-scientific disciplines, such as philosophy. The philosopher Tom Sorell writes that scientism is ‘a matter of putting too high a value on science in comparison with other branches of learning or culture’. But what’s wrong with putting a higher value on science compared with other academic disciplines? What is so bad about scientism? If physics is in fact a better torch in the quest for knowledge than philosophy, as Hawking claimed, then perhaps it should be valued over philosophy and other non-scientific fields of enquiry.

Before we can address these questions, however, we need to get our definitions straight. For, much like other philosophical -isms, ‘scientism’ means different things to different philosophers. Now, the question of whether science is the only way of knowing about reality, or at least better than non-scientific ways of knowing, is an epistemological question. Construed as an epistemological thesis, then, scientism can be broadly understood as either the view that scientific knowledge is the only form of knowledge we have, or the view that scientific knowledge is the best form of knowledge we have. But scientism comes in other varieties as well, including methodological and metaphysical ones. As a methodological thesis, scientism is either the view that scientific methods are the only ways of knowing about reality we have, or the view that scientific methods are the best ways of knowing about reality we have. And, construed as a metaphysical thesis, scientism is either the view that science is our only guide to what exists, or the view that science is our best guide to what exists.

Without a clear understanding of the aforementioned varieties of scientism, philosophical parties to the scientism debate are at risk of merely talking past each other. That is, some defenders of scientism might be arguing for weaker varieties of scientism, in terms of scientific knowledge or methods being the best ones, while their opponents interpret them as arguing for stronger varieties of scientism, in terms of scientific knowledge or methods being the only ones. My own position, for example, is a weak variety of scientism. In my paper ‘What’s So Bad about Scientism?’ (2017), I defend scientism as an epistemological thesis, which I call ‘Weak Scientism’. This is the view that scientific knowledge is the best form of knowledge we have (as opposed to ‘Strong Scientism’, which is the view that scientific knowledge is the only knowledge we have).

According to Weak Scientism, while non-scientific disciplines such as philosophy do produce knowledge, scientific disciplines such as physics produce knowledge that is superior – both quantitatively and qualitatively – to non-scientific knowledge. It is important to note that ‘knowledge’ does not refer to justified true belief (or any other analysis of knowledge, for that matter). Rather, ‘knowledge’ means disciplinary knowledge or the research produced by practitioners in an academic field of enquiry. All academic disciplines are in the business of producing knowledge (or research) in this sense. The knowledge of each academic discipline is what we find in the academic publications of the practitioners of an academic discipline. Proponents of Strong Scientism would deny that non-scientific disciplines produce ‘real knowledge’, as Richard Williams puts it in the introduction to the anthology Scientism: The New Orthodoxy (2014), whereas proponents of Weak Scientism would grant that non-scientific disciplines produce knowledge but argue that scientific knowledge is better than non-scientific knowledge along several dimensions.

Now, whether any of these epistemological, methodological and metaphysical theses – weak (‘best’) or strong (‘only’) – is true should be a matter of debate, not definition by fiat. Each claim needs to be put forward, examined, criticised and debated. It can’t be what some have unfortunately sought to do, which is to make scientism a misguided view by definition. Take the psychologist Steve Taylor who wrote in 2019:

One of the characteristics of dogmatic belief systems is that their adherents accept assumptions as proven facts. This is certainly true of scientism. For example, it is a fact that consciousness exists, and that it is associated with neurological activity. But the assumption that consciousness is produced by neurological activity is questionable.

Here, Taylor asserts that scientism is a dogmatic belief system. But why is that? None of the epistemological, methodological and metaphysical theses mentioned above is dogmatic. These theses can be questioned, of course. However, if merely being questionable were sufficient to make a belief dogmatic, then many if not most of our beliefs would be dogmatic. To see why, consider my (and, in all likelihood, your) belief that there is an external world, a world that is there independently of our minds. Our belief in the existence of an external world is notoriously difficult to prove, as any epistemologist will tell you, but that doesn’t mean that our belief in an external world is nothing more than mere (religious) dogma.

Likewise, in her book Defending Science – Within Reason (2003), Susan Haack asserts that, by definition, scientism is ‘an exaggerated kind of deference towards science, an excessive readiness to accept as authoritative any claim made by the sciences, and to dismiss any kind of criticism of science or its practitioners as anti-scientific prejudice’. But this runs into the same problem above. None of the epistemological, methodological and metaphysical theses mentioned above are exaggerated or excessive. To have an exaggerated deference toward something is misguided, and to have an excessive readiness to accept as authoritative any claims made by some source is foolhardy. After all, that’s just what the words ‘exaggerated’ and ‘excessive’ imply.

To assert that scientism is merely a dogmatic belief, as Taylor does, or an inherently misguided attitude, as Haack does, is to weaponise it, not to argue against it. There is an important history here. In the mid-20th century, theologians and religious scholars weaponised scientism in an attempt to defend their academic territory from what they perceived as a threat of scientific encroachment. In their book Roadblocks to Faith (1954) James Pike and John McGill Krumm distinguished science from scientism and claim that the latter is ‘a threat to the humanities no less than to religion’. Around the same time, in his paper ‘The Preacher Talks to the Man of Science’ (1954), H Richard Rasmusson characterised scientism as ‘a cult that has made a religion out of science’. These religious scholars weaponised scientism out of a concern that science is encroaching on areas of enquiry that presuppose the existence of the very things whose existence they take science to be questioning or denying, such as God, the supernatural, and the like. That is why, according to Ian Barbour, ‘it is [considered] scientism when Richard Dawkins says that the presence of chance in evolution shows that this is a purposeless universe,’ for this claim is taken to be questioning the belief in a providential God.

Some philosophers are now playing a similar game, that is, using scientism as a weapon in the fight against scientists who are critical of academic philosophy. Philosophers who level the charge of ‘scientism’ typically identify prominent scientists, such as Hawking and Neil deGrasse Tyson, as exhibiting this kind of misguided attitude toward science; the philosopher Ian James Kidd called them mere ‘cheerleaders for science’. The problem with thinking of scientism as these philosophers do – as exaggerated deference toward science – is that it is a persuasive conception of scientism. To assert that scientism is ‘putting too high a value on science’ or ‘an exaggerated kind of deference towards science’ is to express disapproval of what could, after all, be a reasonable view to hold.

In argumentation studies, definitions that are intended to transfer emotive force, such as feelings of approval or disapproval, are known as persuasive definitions. To say that scientism just is ‘putting too high a value on science’ is akin to saying that abortion is murder – the definition is overloaded with emotional force. Just as pro-choice advocates would object to saying that abortion is murder, since it expresses disapproval of abortion, advocates of scientism would object to saying that scientism is ‘putting too high a value on science’ since it expresses disapproval of scientism. Instead of condemning scientism by definition, as Sorell and Haack do, opponents of scientism need to show precisely what is wrong with it.

In their introduction to the anthology Scientism: Prospects and Problems (2018), René van Woudenberg, Rik Peels and Jeroen de Ridder agree that scientism should not be weaponised when they write: ‘no one will accept this notion of “scientism” as an adequate characterisation of their own views, as no one will think that their deference to science is exaggerated, or their readiness to accept claims made by the sciences is excessive.’ In fact, in the paper ‘Six Signs of Scientism’ (2012), Haack herself observes that, before it was weaponised by those who sought to defend religion and philosophy from science trespassing on their territories, ‘the word “scientism” was neutral.’

Unlike pejorative conceptions of scientism, Weak Scientism – the view I defend – is a neutral framing according to which scientific disciplines, when compared with non-scientific disciplines, such as philosophy, are better along several dimensions. Briefly, the argument runs as follows. One thing can be said to be better than another thing either quantitatively or qualitatively. Scientific knowledge is quantitatively better than non-scientific knowledge because scientific disciplines produce more knowledge, and the knowledge they produce has more impact than the knowledge produced by non-scientific disciplines. This claim is supported by data on the research output (that is, number of publications) and research impact (that is, number of citations) of scientific and non-scientific academic disciplines. These data show that scientific disciplines produce more publications, and those publications get cited more than the publications of non-scientific disciplines.

In their paper ‘Humanities: The Outlier of Research Assessment’ (2020), Güleda Doğan and Zehra Taşkın use publication and citation data in 255 subjects on the Web of Science from 1980 to 2020 to find that the distribution of publications is such that ‘81 per cent were published in three main pure sciences categories: natural sciences (33 per cent), medical sciences (27 per cent), and engineering and technology (21 per cent)’, and that ‘[t]he total number of humanities publications was almost similar to a relatively small pure science area’, namely, agricultural sciences. As far as the distribution of citations is concerned, the ‘[h]umanities had only 0.52 per cent of whole citations in the dataset, while natural sciences had 44 per cent, medical sciences had 30 per cent, engineering and technology had 17 per cent, social sciences had 6 per cent, and agriculture had 1.5 per cent.’ When compared with the natural sciences, engineering and technology, medical and health sciences, and social sciences, the humanities have the lowest values of research output (measured by publication counts), with the exception of agricultural sciences, and research impact (measured by citation counts), without exception. While most scientific publications are cited, only 16 per cent of publications in the humanities are cited. Within the humanities, philosophy, ethics and religion have the highest percentages of uncited publications.

Scientific knowledge can be said to be qualitatively better than non-scientific knowledge because scientific knowledge is explanatorily, predictively and instrumentally more successful than non-scientific knowledge. This is the sort of success that philosophers of science talk about, as they often do, when they say that science is successful. Consider, for example, Albert Einstein’s theory of relativity. The theory is explanatorily successful insofar as it provides a comprehensive explanation for phenomena that would otherwise seem mysterious, such as gravity, planetary orbits, black holes, electromagnetism, and more. The theory is instrumentally successful insofar as it allows us to intervene in nature as when we use GPS to navigate our world and gravitational lensing to look for new worlds. The theory is predictively successful insofar as it makes novel predictions that are borne out by observation or experimentation, such as the perihelion precession of Mercury, the deflection of light by massive objects, the gravitational redshifting of light, the relativistic delay of light (also known as the Shapiro effect), gravitational waves, and more. One would be hard pressed to find a non-scientific theory that is as explanatorily, instrumentally and predictively successful as the theory of relativity.