[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123503-en":3,"doc-seo-123503-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},123503,1099514068365,"Aurelia","https://ap-avatar.wpscdn.com/avatar/10000253d8d9f28188e?_k=1776742907772140068",8,"Research & Report","MACHINE LEARNING AND THEORY-LADENNESS - A PHENOMENOLOGICAL ACCOUNT","This article analyzes theory-ladenness in machine learning within science, treating “theory” as domain knowledge (domain-theory) from the discipline where ML is applied. By comparing machine-learning model construction with phenomenological models, it argues that ML model-building is largely indifferent to domain-theory. Even so, models remain theory-laden in a weakened form termed “theory-infection.” The analysis highlights consequences for transferring ML across scientific disciplines and shifts the debate from descriptive to normative.","MACHINE LEARNING AND THEORY-LADENNESS: A  \nPHENOMENOLOGICAL ACCOUNT  \nAlberto Termine, IDSIA USI-SUPSI, Lugano  \nEmanuele Ratti 1, University of Bristol  \nAlessandro Facchini, IDSIA USI-SUPSI, Lugano  \nAbstract. We provide an analysis of theory-ladenness in machine learning (ML) in science, where‘theory’ (that we call 'domain-theory') refers to the domain knowledge of the scientific discipline where ML is used. By constructing an account of ML models based on a comparison with phenomenological models, we show (against recent trends in philosophy of science) that ML model-building is mostly indifferent to domain-theory, even if the model remains theory-laden in a weak sense, which we call theory-infection. These claims, we argue, have far-reaching consequences for the transferability of ML across scientific disciplines, and shift the priorities of the debate on theory-ladenness in ML from descriptive to normative.  \n1. INTRODUCTION  \nThe development of data-intensive methods in the sciences (from Big Data, data science, to AI) has been often associated with the idea that these can function without inputs from scientific expertise, or without appealing to domain knowledge of the scientific fields in which they are used. This idea was especially central in the debate on the so-called ‘two cultures of statistical modeling’(Breiman 2001) . Famously, Breiman distinguishes two ways of doing statistics, one more “traditional”(called ‘data modeling culture’) based on the adoption of socalled data-models that statisticians formulate with the help of domain-related theoretical considerations, and one relying on far less assumptions and based on learning predictive algorithmic models directly from data with few constraints (called ‘algorithmic modeling’) . The way Breiman describes the so-called ‘algorithmic modeling culture’ overlaps significantly with the strategies associated with concepts like ‘data-intensive scientific discovery’(Hey et al 2009),‘data-driven’ science, Big Data, machine learning, and now, more in general, AI2. In all these cases, there is an emphasis on leveraging data by relying less on domain assumptions, and more on a few constraints given by ‘general-purpose’ tools that can be easily adapted to a variety of application domains and tasks. The prototypical example of this tendency are arguably multi-modal foundation models, which are increasingly common in contemporary AIresearch. These are large deep neural networks pre-trained on huge corpora of data (including  \n1 Corresponding author, [mnl.ratti@gmail.com](mnl.ratti@gmail.com)  \n2 Such independence has been even popularized as a new scientific paradigm which came to be known, infamously, as the‘End of Theory’ (Anderson 2008)  \nimages, texts, audio etc.), and conceived to be easily adaptable to a variety of tasks via a simple fine-tuning or even with just a proper prompt through an API. .  \nGauging whether data-intensive methods are independent from, or less reliant on, theoretical considerations coming from the scientific domain of implementation means asking a question about the theory-ladenness of data-intensive methods. By ‘theory-ladenness’, we mean here the idea that, in data-intensive science, engaging in essential scientific activities requires either the implicit assumption of, or the explicit appeal to, scientific theories. We understand the term‘theory’ in a broad sense (specified in more detail below) to include domain knowledge and expertise of the given scientific field in which data-intensive methods are used. While in data-intensive science - and especially in machine learning (ML) -there might be ideas coming from statistical learning theory that can reasonably count as ‘theory’, here theory-ladenness is restricted to the ‘domain-theory’ belonging to the scientific context of implementation of the model.  \nWith few exceptions (e.g., Napoletani et al. 2020), philosophers of science have been systematically arguing in favor of theory-ladenness (Cal","cbCaiaQfumH5bwSp","https://ap.wps.com/l/cbCaiaQfumH5bwSp","pdf",583327,1,38,"English","en",105,"# Introduction\n## Two cultures of statistical modeling\n## Defining theory-ladenness and domain-theory\n## Philosophical background: arguments for theory-ladenness\n## Aim and core thesis for ML model-building","[{\"question\":\"How does the article define “theory-ladenness” in data-intensive science and machine learning?\",\"answer\":\"It defines theory-ladenness as the idea that essential scientific activities in data-intensive science require implicit assumptions or explicit appeals to scientific theories. In this work, “theory” is treated broadly to include the domain knowledge used in the model’s scientific context.\"},{\"question\":\"What is the article’s main claim about domain-theory in machine-learning model building?\",\"answer\":\"The article argues that practices involved in constructing machine-learning models are mostly theory-indifferent, meaning they do not necessarily require explicit reference to the domain-theory.\"},{\"question\":\"What are the consequences of theory-indifference for transferring ML across scientific disciplines?\",\"answer\":\"If ML model-building is theory-indifferent while models remain weakly theory-laden through “theory-infection,” the work argues this affects transferability across disciplines. It also reframes the debate on theory-ladenness in ML toward normative priorities.\"}]","MACHINE LEARNING AND THEORY-LADENNESS - A PHENOMENOLOGICAL ACCOUNT | PDF",1785816896,96,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"machine-learning-and-theory-ladenness-a-phenomenological-account-123503","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/machine-learning-and-theory-ladenness-a-phenomenological-account-123503/123503/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-04",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"How does the article define “theory-ladenness” in data-intensive science and machine learning?","Question",{"text":75,"@type":76},"It defines theory-ladenness as the idea that essential scientific activities in data-intensive science require implicit assumptions or explicit appeals to scientific theories. In this work, “theory” is treated broadly to include the domain knowledge used in the model’s scientific context.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What is the article’s main claim about domain-theory in machine-learning model building?",{"text":80,"@type":76},"The article argues that practices involved in constructing machine-learning models are mostly theory-indifferent, meaning they do not necessarily require explicit reference to the domain-theory.",{"name":82,"@type":73,"acceptedAnswer":83},"What are the consequences of theory-indifference for transferring ML across scientific disciplines?",{"text":84,"@type":76},"If ML model-building is theory-indifferent while models remain weakly theory-laden through “theory-infection,” the work argues this affects transferability across disciplines. It also reframes the debate on theory-ladenness in ML toward normative priorities.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]