NOTE-107: Religion and Science (SEP)

Contribution

The entry is an orientation to the interdisciplinary field of “science and religion” (also “theology and science”): its institutions, history, key taxonomies and its extension beyond Christianity. It adds no new thesis. What it contributes is a synthesis in which the relationship varies by tradition and by specific claim, and in which the global conflict-versus-harmony framing is historically shallow.

Key insight

“Science” and “religion” are both nineteenth-century comparative categories (§1.2). Whether they conflict is therefore not one question. It depends on the tradition, which creedal tenets are in play, and the history of contact, including colonialism. So the productive unit of analysis is a specific scientific claim paired with a specific theological claim.

Assumptions

Premises and authorities, since this is a survey.

• Scope (§1.1): the entry treats science and religion “as rather abstract philosophical positions, rather than as practices”, and defers the sociology to Ecklund and others.
• Historically, the field has been “de facto” Western science plus Christianity (§1.1). Section 3 keeps a Christian focus “due to its prevalence in the literature”.
• Main authorities: Barbour (1966, 2000) for the taxonomy; Harrison (1998, 2007, 2015) and Bowler (2001, 2009) for history; Ungureanu (2019) on the conflict thesis; Lopez (2008) and McMahan (2009) on Buddhist modernism; Guessoum (2011) and Edis (2007) on Islam; C. M. Brown (2008, 2012) on Hinduism; Peters and Hewlett (2003) for the divine-action spectrum; Perry and Ritchie (2018) for “science engaged theology”.

Key results

What the entry reports and argues, attributed to it unless it attributes onward.

• History (§1.1). The systematic field dates from the 1960s: Barbour 1966, Torrance 1969, and the journal Zygon founded in 1966. Harrison (1998) argued Protestant conceptions helped give rise to seventeenth-century science. The Vatican Observatory and CTNS divine-action project of the 1990s produced six volumes. In the public sphere the conflict view prevails, “in stark contrast to the scholarly literature”, which the entry attributes in part to a vocal conservative Christian minority in the US (Evans 2016; §2.1).
• Demarcation and naturalism (§1.2). Methodological naturalism is an epistemological principle limiting science to natural entities and laws. Ontological naturalism is a metaphysical principle rejecting the supernatural (Forrest 2000). Demarcation by methodological naturalism “is more commonly accepted”, illustrated by Pennock’s testimony in Kitzmiller v. Dover (2005). Draper (2005) argues the success of methodological naturalism may be evidence for ontological naturalism. Rosenberg (2014) draws stronger conclusions, and the entry calls these “controversial”.
• Taxonomies (§1.3). Barbour’s (2000) conflict/independence/dialogue/integration is “the most influential”. Stenmark’s (2004) independence/contact/union is an alternative. The entry’s own criticisms: the categories are “not clearly defined and … not mutually exclusive” (Tennant 1902 is both conflict and integration); integration is “skewed towards theism”; and Gould’s NOMA faces the problem that religions make empirical claims and that values need factual grounding (Worrall 2004). Cantor and Kenny (2001) are cited as doubting the taxonomy’s historical usefulness.
• Scientific study of religion (§1.4). The arc runs from natural histories of religion (Fontenelle 1724, Hume 1757) through cultural evolutionism, secularization theses and functionalism, to fieldwork (Evans-Pritchard’s Azande, who combine natural and witchcraft explanations) and the cognitive science of religion. CSR presents religion as a byproduct of ordinary cognition (Bloom 2007) or as an adaptation for cooperation (Norenzayan 2013). The entry calls the order of “big god beliefs or large-scale societies” “a continued matter of debate”. Legare et al. (2012) are cited for the cross-cultural mixing of natural and supernatural explanations.
• Five traditions (§2). Christianity: the two-books metaphor; concordism, which Lamoureux (2008) rejects because scientific-sounding biblical statements are false; critical realism as the dominant epistemology; Harrison (2007) on original sin motivating instrument-based science. Stark’s (2004) Christian-uniqueness thesis is criticised for neglecting Islamic and Greek contributions. Islam: science flourished c. 9th–15th century under Abbasid patronage and later stagnated for reasons the entry calls “unclear”. It prefers Griffel’s and Edis’s emphasis on legal changes (the widening of zandaqa) and the colonial association of science over the “al-Ghazālī killed science” narrative. Old Earth creationism dominates. Guessoum (2011) upholds Ibn Rušd’s no-possible-conflict principle. Hinduism: there is no sharp creator/creature distinction, so ontology differs. Vedic ritual drove astronomy, geometry and positional notation. Responses to evolution ranged from Dayananda’s rejection to Vivekananda’s and Aurobindo’s integration. Buddhism: modernist compatibility claims are qualified by Chinese and Japanese Buddhists’ rejection of the struggle for existence in favour of Kropotkin’s mutual aid (Ritzinger 2013; Burenina 2020). Judaism: Rabbinic non-literal reading leaves room for science. Cohen’s neo-Kantian complementarity, Kook’s NOMA-like view and Kaplan’s dialogue and process theology are surveyed, along with the 1885 Pittsburgh Platform.
• Divine action (§3.1). The doctrine of creation has four features: ex nihilo, God distinct from the world, the goodness of creation, and eschatology. It distinguishes general from special divine action (Wildman 2008) and direct from indirect acts (Alston 1989). The divine-action spectrum (Peters and Hewlett 2003) runs from creationism, through Intelligent Design, theistic evolution and deism, to ontological materialism. Interventionism (Hume’s and Swinburne’s miracle-as-violation) presupposes determinism. Non-interventionism needs indeterminism somewhere, and the candidates are chaos (Polkinghorne) and quantum events (Russell, Murphy). These are criticised as “physicalism-plus-God” (Jaeger) and as a category mistake (Carroll’s neo-Thomism: God as primary cause grounds secondary causes and needs no indeterminacy). Alston (1989) argues a deterministic world is compatible with divine action, since every event is then an indirect divine act. On chance and providence, Plantinga (2011) says God may guide mutations, while Johnson (1996) makes randomness a secondary cause. The free-will and “only way” theodicies are discussed (Miller 1999; Southgate 2008).
• Human origins (§3.2). It surveys the hominin record: Homo sapiens “as early as 315,000 years ago”, Denisovans, a branching tree. Theological responses on uniqueness include van Huyssteen, and extending the imago Dei to animals (Deane-Drummond, Putz). It covers Teilhard’s progressivism and Deane-Drummond’s critique, and on the historicity of Adam, van Inwagen and J. K. A. Smith against Schneider.
• Close. The literature “has split up in diverse fields of ‘science engaged theology’” (Perry and Ritchie 2018), in which a specific scientific claim is paired with a specific theological one.

Claims

id	claim	strength	support
C1	”Science” and “religion” in their current senses are nineteenth-century coinages	moderate	secondary historical sources cited (Harrison 2015; J. Smith 1998; Whewell 1834)
C2	Most scholars in the field reject the conflict model as based on “a shallow and partisan reading of the historical record”	assertion	the author’s characterisation of the field; no survey cited
C3	The conflict thesis was not originally a secularist project; Draper’s and White’s aims were liberal-Protestant	moderate	a single cited source (Ungureanu 2019)
C4	Barbour’s categories are neither exhaustive nor mutually exclusive	informal argument	the Tennant counterexample and McGrath’s POMA as a hybrid
C5	Methodological naturalism is the more commonly accepted demarcation of science from religion	assertion + legal record	the Kitzmiller testimony cited as illustration
C6	Legal changes and colonial associations did more to stifle science in the Islamic world than al-Ghazālī did	moderate	secondary sources (Edis 2007; Griffel 2009); the entry says the ultimate cause “remains unclear”
C7	Quantum or chaos loci for non-interventionist divine action face serious objections	informal argument (reported)	Jaeger 2012a,b; Saunders 2002; Carroll 2008; the entry adds the epistemic/ontic slide in Polkinghorne
C8	Buddhist responses to evolution anticipated the recognition of cooperation as an evolutionary principle	weak	an interpretive claim built on Ritzinger 2013 and Burenina 2020
C9	The field is fragmenting into claim-specific “science engaged theology”	assertion	Perry and Ritchie 2018 cited

Concepts

• Methodological vs ontological naturalism. The first limits scientific inquiry to natural entities and laws and is silent on whether the supernatural exists. The second denies the supernatural (§1.2, after Forrest 2000).
• Barbour’s taxonomy. Conflict, independence, dialogue, integration. Integration comes as natural theology, theology of nature, or process philosophy (§1.3).
• NOMA / POMA. Gould’s non-overlapping magisteria (facts vs values and meaning), and McGrath’s partially overlapping magisteria.
• General vs special divine action. Creation and sustenance, as against particular providential acts (Wildman 2008). Direct vs indirect. Acting without or through natural causes (Alston 1989).
• Interventionism vs non-interventionism. Whether special divine action suspends laws, which presupposes determinism, or works within indeterministic gaps.
• Primary vs secondary causation. The neo-Thomist view that God grounds creaturely causes rather than competing with them (Carroll 2008; Johnson 1996).
• Science engaged theology. Pairing a specific scientific claim with a specific theological one (Perry and Ritchie 2018).

Connections

The metaphysical theses the entry reports rather than defends are ontological naturalism versus supernaturalism; determinism and whether physical laws leave causal room for non-physical agency; and the primary/secondary causation model, on which a transcendent cause does not compete with natural causes. Its own stance presupposes only that these are live questions and that methodological naturalism characterises science. The §3.1 structure has the same shape as the causal-exclusion problem, including a “category mistake” reply that denies the causes compete. That parallels the downward-causation debates held elsewhere in the record. LIT-192 (on naturalizing teleology) shares the §1.2 question of what naturalism demands of metaphysics. LIT-129 (epistemic injustice toward practitioners of Afro-Brazilian religions) instantiates §1.4’s point, via Legare et al., that people combine natural and supernatural explanations. LIT-066 (meditation-related challenges in Western Buddhists) bears on §2.4’s report of Buddhist-modernist claims that mindfulness is corroborated by science. None of these is cited by the entry.

metaphysics tag: justified, as a secondary tag. Laws, causation and determinism in §3.1, and ontological naturalism in §1.2, fall under the blurb. Most of the entry is history and survey.

Bearing on the record

It carries nothing for ML practice, and saying so is the answer. There is no instruction for the Anthology of the SOTA and no THEORY document here to support or contradict. Its main use to this record is as a reference map for the naturalism and divine-action literature if the record takes that up. It is also a data point that the topic vocabulary cannot yet say “religion”.

Note of 2026-10-16: the vocabulary can now say it. ADR-007 added the religion topic, and LIT-194 is tagged with it first.

Limitations

• It is a survey. Characterisations of scholarly consensus (C2, C5) are the author’s, with no systematic evidence.
• Its centre of gravity is Christian (§3 explicitly, and §2.1 is the longest subsection). The non-Christian traditions each get a few paragraphs, and some claims about them are compressed; for example, Islamic scientific stagnation is described as unexplained.
• It treats science and religion as ideas, not practices, so the sociological literature is only pointed to (§1.1).
• The quantum and chaos material is not engaged with any particular interpretation of quantum mechanics beyond naming Copenhagen.
• The typescript shows minor slips, such as “Morcedai Kaplan” and “Nishiren”. “The relationship between Buddhism and religion” (§2.4) should presumably be “Buddhism and science”.

Open questions

• Whether a replacement for Barbour’s taxonomy exists that sorts positions exclusively. The entry shows the old one fails but offers none.
• Whether any non-interventionist model of divine action survives the combined “physicalism-plus-God” and category-mistake objections. The entry reports both sides without adjudicating.
• Whether the science–religion relation in non-Christian traditions has its own general structure, or only case histories. The entry gives only the latter.

Corrections to the seeded skim

• Version: the text read is the substantive revision of 3 Sep 2022, not the 2017 first publication. The LIT’s published: 2017-01-17 is right for the entry, but any quotation should be dated to the 2022 revision.
• Dossier on §1.3: “the conflict thesis did not in fact originate with Draper and White”. The entry is subtler. It still calls Draper (1874) and White (1896) “early architects of the conflict thesis”. What it denies, following Ungureanu (2019), is that their project was secularist: it was liberal Protestant, meant to separate religion from theology and so save religion, and was only later appropriated by sceptics.
• Dossier: “the close recommends ‘science-engaged theology’ (Perry and Ritchie 2018)”. The close reports that the literature “has split up in diverse fields of ‘science engaged theology’”. The nearest thing to a recommendation is §1.1’s remark that “productive questions focus on specific topics”. The LIT summary’s “the field is moving” is supported; “recommends” overstates.
• The dossier’s check “whether §3.1 engages with interpretations of quantum mechanics as loci for non-interventionist divine action” can be answered yes, briefly. It covers Russell (2006), God acting in quantum events under the Copenhagen interpretation; Murphy (1995), a bottom-up model in the space of quantum indeterminacy; and Polkinghorne (1998), “ontological openness” from chaos. It reports the criticism as well: Jaeger’s (2012a) “physicalism-plus-God” and Carroll’s (2008) neo-Thomist charge of a category mistake. It is one paragraph and does not compare interpretations, so the quantum-foundations link stays thin, and I would not add the tag.
• The subject is the relation of science to religion, and the closed vocabulary has no word for philosophy of religion or religion. philosophy-of-science and metaphysics are the nearest honest tags, but the record should consider a topic such as religion (or philosophy-of-religion), with a decision, if it files more of this literature. LIT-129 (clinical care of practitioners of Afro-Brazilian religions) would be a second holder.
• Optional additional tag: social-science. §1.4 is a history of the anthropology, sociology and psychology of religion (Tylor, Durkheim, Evans-Pritchard, Allport and Ross, cognitive science of religion, Norenzayan), and someone browsing social science would be right to expect it.
• metaphysics tag: justified, as a secondary tag. The §1.2 and §3.1 material is metaphysics in the blurb’s sense (laws, causation, determinism, naturalism), but it is a minority of the entry. The primary topic, philosophy-of-science, should stay unless a religion topic is added, in which case that should lead.