About the Project

The following summary is based on the original research proposal.

Significance of the Research

(1) Historical Crisis and Digital Opportunity

The humanities, particularly the foundational disciplines of philology and classical studies, are currently facing a crisis. In an era that strongly demands immediate utility—prioritizing large-scale, interdisciplinary, short-term, and results-oriented research—studies that require continuous critical analysis of materials over spans of a century or more are often undervalued. The gradual elucidation of specialized topics is seen as disproportionately labor-intensive with meager visible outputs, appearing “inefficient” by modern quantitative standards. Furthermore, the knowledge required for such research is becoming increasingly complex and growing exponentially over time, making it difficult for individual researchers to rely solely on their personal expertise.

Despite these challenges, humanity continues to seek meaning by referencing the sacred texts, classics, and history inherited from the past. However, as current global tensions demonstrate, leaving the interpretation of ethnic, religious, and historical meanings to arbitrary judgment can endanger both humanity and the planet. Therefore, the critical study of these intellectual legacies—entrusting the interpretation of sacred texts and classics to reason—is an endeavor of human wisdom that must be steadfastly maintained and passed on.

A unique opportunity to overcome this crisis in the humanities has arrived: the Information and Communications Technology (ICT) revolution. Since the 1990s, this phenomenon has transformed the entire research process—from data collection and analysis to the publication and sharing of results—into a digital scholarly space capable of processing information with exponentially higher efficiency. The humanities can solve the aforementioned challenges by creating this new academic space and realizing a Digital Transformation (DX) that re-establishes the discipline within it. For philology and classical studies, the most crucial task is textual criticism—the act of constructing a reliable foundation for interpretation. This prerequisite, which determines the quality of interpretation itself, can only be fulfilled by exploring the intellectual heritage of different cultures and worlds as broadly and deeply as possible and subjecting the results to critical scrutiny. This endeavor, once becoming impossible within the closed knowledge space of print media, has now gained unprecedented possibilities through the power of Artificial Intelligence (AI) in digital media. However, the Japanese humanities have yet to fully recognize this reality, responding only with ad-hoc measures and failing to seize this opportunity.

(2) Project Objectives and the “Reiwa Daizōkyō”

To fundamentally resolve this issue, this project focuses on Buddhist Studies, the field within the Japanese humanities where the digitization of the knowledge base is most advanced. Utilizing the “Next-Generation Humanities DX Model” and a newly developed “Classical Text Compilation System” equipped with AI-OCR (both developed by the Principal Investigator), we will undertake a large-scale DX to enable the re-compilation of the entire Buddhist Canon—specifically, the compilation of the “Reiwa Daizōkyō”. Through this process, we will present insights into new research methodologies and academic evaluation standards, demonstrating the path forward for Japanese humanities in the DX-AI era. This is a pioneering attempt to open up a new horizon of scholarship never before achieved in any field of the Japanese humanities.

The key points of this project are outlined below in three sections:

1. Construction of the “Collaborative Research Database for the Humanities (CRDIH)” for re-editing the Canon on the web.
2. Differentiation from other similar projects.
3. Establishment of a new academic research process and evaluation method in the age of Open Science.

1. Construction of the “Collaborative Research Database for the Humanities (CRDIH)”

The Taishō Shinshū Daizōkyō (Taishō Canon), which presented the Buddhist corpus formed over 1,500 years to the modern world, has been accepted as a standard reference book in research libraries globally. It has played a unique and irreplaceable role as the knowledge foundation for Eastern philosophy, literature, and history. This project creates a new academic environment where research on this sacred corpus can utilize the superior capabilities of ICT, including AI. Specifically, by using the latest AI-OCR technology, we will enable character-by-character comparison of over 100 million characters in the Canon across various editions (Song, Yuan, Ming, Goryeo, and the Imperial Household Agency Song versions). Through this, we will attempt to revise the Canon corpus in a collaborative web-based workspace.

This attempt to reconstruct the entire developmental process of a massive sacred canon is a visualization of the knowledge construction process on a grand scale, something previously impossible. This will establish a new research infrastructure for textual criticism, aggregating vast knowledge accumulated over the past century, and realize the academic community’s century-long dream: the re-compilation of the Buddhist Canon. It is a historic undertaking that refreshes Japan’s contribution to the international academic community for the next hundred years.

2. Differentiation and Positioning of the Research

Japan possesses technology and achievements comparable to advanced nations in digitizing and releasing materials. However, in terms of Humanities DX—incorporating unique humanities insights and sharing them in a usable form leveraging digital characteristics—Japan lags significantly behind. This project presents a powerful model to solve the challenges Japanese humanities will face. The success of the CRDIH proposed by this project concerns not only Buddhist studies but the very survival of Japanese humanities in the digital age. For large-scale research institutions to fully unleash their potential, research that challenging pioneers new fields is essential. This project aims to fulfill that role from the field of Buddhist studies, and its realization will greatly benefit the Japanese humanities as a whole.

3. New Research Process and Evaluation in Open Science

Despite the Japanese government allocating vast resources to Open Science, the humanities have not sufficiently responded. The CRDIH research platform aims to open a concrete path for the humanities to contribute to this challenge.

• Contributors: Researchers involved in building the CRDIH will have their contributions recorded by name and will be granted the right to claim academic credit.
• Users: Researchers who use the revised results in their work and provide academic judgment on their selection will similarly have their input recorded in the CRDIH and receive academic credit.
• Dynamic Interaction: In highly academic exchanges, providers and users can swap roles, building a platform where the response itself is recognized as a research result.

The CRDIH addresses two challenges: the traditional task of judging highly specialized academic quality in Buddhist studies, and the new, common challenge of establishing academic evaluation methods on a web-based platform. The CRDIH Editorial Board and the Advisory Board, composed of experts in Buddhist studies and leading Digital Humanities scholars from the West, will guide these efforts. We will explore an academic evaluation method that introduces the concept of “Preprints”—enabling speed and diversity—into traditional authorization methods, allowing both to coexist in a layered manner. Through this activity, a path will be opened for the humanities to respond to the challenges posed by Open Access.

Japan faces the critical task of inheriting its vast historical and cultural heritage as an intellectual endeavor for the next century and continuing to transmit its significance to the world. This cannot be solved without viewing the entire research process—from initiation to execution and evaluation. This project aims to squarely take on this challenge as a single academic discipline and present a concrete model for building the next-generation humanities research infrastructure.

1．Academic Background and Key Research Questions

Historical Background: The Legacy of the Buddhist Canon

Buddhism, which emerged in India in the 5th century BCE, produced a vast amount of sacred texts that spread across Asia, becoming a pillar of Eastern thought and culture. Among these, the scriptures that were intermittently introduced and translated into Chinese over more than a millennium formed a magnificent collection of knowledge comprising over 5,000 volumes known as the Daizōkyō (The Great Buddhist Canon). Since the Catalogue of Scriptures (Zongli Zhongjing Mulu) by Shi Daoan in the 4th century, the Daizōkyō has been continuously compiled and reorganized within the East Asian intellectual tradition. From the 10th century onwards, it was transmitted through woodblock printing in China, Korea, and Japan.

In the late 19th century, amidst the various branches of the Daizōkyō that had developed over time, Shimada Bankon and others in Japan published the Dainippon Kōtei Daizōkyō (also known as the Shukusatsu-zō). This was the world’s first Buddhist canon published using movable type, based on critical collation with three major editions (the Goryeo Re-carved Edition, the Song Edition, and the Yuan Edition) held at Zōjōji Temple in Tokyo. Building on this foundation and incorporating new comparative materials, Takakusu Junjirō and his colleagues re-edited and published the Taishō Shinshū Daizōkyō (hereafter “Taishō Canon”) as an international series starting in 1924. The Taishō Canon has been accepted as a standard reference book in research libraries worldwide due to its rarity and high academic reliability. It has played a unique and irreplaceable role as a knowledge foundation for research on Eastern thought, literature, and history emanating from Japan.

The Digital Turn: Establishing the SAT DB as a New Commons

With the rapid progress of the Information and Communications Technology revolution since the late 20th century, the academic world began to experience a massive transformation. The analysis of this unprecedented challenge—reorganizing the entire research process from data collection to publication into a digital scholarly space—revealed a crucial fact: the Taishō Canon has functioned as a network, or a “Buddhist Studies Commons,” enabling the formation of an international community of scholars and the creation of knowledge.

The SAT Daizōkyō Text Database Study Group (Representative: Masahiro Shimoda; hereafter “SAT”), established in 1994, aimed to realize the significance and utility of this “Commons” in the digital space. By constructing the SAT DB, a full-text database of the Taishō Canon comprising over 100 million characters, we sought to create a new academic infrastructure for Buddhist studies and related disciplines. Within this grand endeavor lies the core academic question of this research project.

International Collaboration and Standardization Efforts

In reorganizing this academic infrastructure, the Principal Investigator and the team proceeded with the research and construction of the SAT DB while forming a new international digital research network. In this process, we recognized that the Digital Transformation (DX) of Japanese humanities was lagging, leading to isolation from international networks. We realized that this isolation would become a major obstacle to the future sustainability of the SAT DB. In the digital academic environment, diverse knowledge must be networked in a unified format, transcending differences in language, institutions, disciplines, nations, and regions. Improving Japan’s digital environment is, therefore, a condition for the sustainable development of SAT.

To achieve this, in cooperation with like-minded domestic researchers, the Principal Investigator served as the founding president of the Japanese Association for Digital Humanities (JADH) in 2012, an allied organization of the Alliance of Digital Humanities Organizations (ADHO). Since then, we have held annual international conferences, inviting leading DH researchers from abroad to conduct lectures and tutorials for the Japanese humanities community. Furthermore, we hosted the annual international conference of the TEI Consortium—which formulates the Text Encoding Initiative (TEI) guidelines, the international standard for text data construction in humanities—in Tokyo in 2018, attracting over 300 participants. In 2022, we hosted the ADHO annual conference online in Tokyo, welcoming 700 participants from over 40 countries. Through close collaboration with international DH activities, we have promoted the establishment of international standards and human networks necessary for the digitization of the humanities.

Technical Contributions: Unicode, TEI, and IIIF

To make these activities substantive, an environment capable of appropriately handling East Asian materials, including Japanese, on digital media was necessary. However, international standards defining such an environment remained insufficient for a long time.

• Character Encoding: SAT officially joined ISO/IEC JTC1/SC2, the committee for Unicode standardization, as a liaison member. We successfully registered over 3,000 characters—including Kanji and Siddham (Sanskrit) variants that could not be handled properly on computers—into Unicode via ISO/IEC 10646.
• Text Encoding: While the TEI Guidelines were the de facto international standard, they were Western-centric and did not sufficiently support East Asian text materials. In 2016, SAT members established the East Asian/Japanese Special Interest Group (SIG) as an official body within the TEI Consortium, leading to the revision of the TEI Guidelines in 2021.
• Image Interoperability: Regarding web image processing, which is vital for manuscript and printed book research, we released Japan’s first web service compatible with IIIF (International Image Interoperability Framework). We also improved IIIF image viewers for East Asian literature, advancing the formation of infrastructure for research on Japanese and East Asian culture.

We aggregated these achievements and presented them as the “Model of Digital Research Environment for Next-Generation Humanities” (hereafter “Next-Generation Model”). To improve the digital research environment in Japan, we published open-access books such as How to Create a Digital Scholarly Space (2019) and held over 100 seminars on TEI Guidelines and IIIF at various institutions. Our efforts supported the adoption of IIIF at major institutions like the National Diet Library and the National Institute of Japanese Literature (NIJL), and promoted the use of TEI Guidelines at the University of Tokyo, the National Museum of Japanese History, and others. These activities have been highly acclaimed, leading to keynote lectures at international DH conferences (e.g., Hamburg 2012) and invited lectures worldwide. Domestically, we received the 8th Gesner Award (Gold Prize in the Digital Knowledge Organization category) and the 1st Society Award (Practice Award) from the Japan Society for Digital Archive in 2019.

The Challenge of Generative AI and Future Directions

Recently, the Japanese humanities, which had been gradually resolving the delay in DX, faced a new, significant problem that cannot be addressed by conventional measures: the emergence of Generative AI. It is now an urgent task for all research fields to consider Generative AI, which has a massive impact on the research process and is rapidly becoming a de facto standard research method. The quality of responses generated by AI depends entirely on the existence and quality of training data. In Japan, where the open access and standardization of humanities data have hardly been realized, the amount of reliable training data is scarce. If this situation continues, Generative AI in the Japanese-speaking world will be left behind by the rapidly advancing global standards, and the humanities may fade away without benefiting from it.

Amidst this crisis, SAT succeeded in developing an automatic collation system for woodblock prints and manuscripts utilizing developing AI and AI-OCR technology, dramatically improving the efficiency of processing research materials. The results were reported at international conferences, starting with the University of Vienna in April 2023, and have been continuously improved through reviews. Concurrently, we have begun developing an academic research support system using Generative AI with RAG (Retrieval-Augmented Generation), which was presented at international symposia at UC Berkeley and the University of Tokyo. By placing these systems—reorganized by applying AI to the “Next-Generation Model”—at the core of our research and engaging scholars from related fields, we can now realize the century-long dream of the academic community: the re-compilation of the Taishō Canon. Through this, the Buddhist Studies Commons of the past will be refreshed for the future, illuminating the path forward for the humanities.

2. Research Objectives, Originality, and Creativity

Objectives: Reorganizing the Humanities for the Digital Age

The primary objective of this research is to identify the requirements for the humanities to flourish within a new knowledge environment created by Digital Transformation (DX) and AI, and to construct a digital scholarly space where these requirements are integrated coherently. Ultimately, we aim to present a detailed roadmap for the developmental reorganization of the humanities as a whole. To achieve this, we focus on Buddhist Studies, the field where the digitization of the knowledge base is most advanced. By re-editing the Buddhist Canon (Daizōkyō)—the foundation of knowledge in this field—on an academic scale, we will build a new “Researcher Commons.” Realizing this formation of scholarly knowledge in digital media requires constructing data using a unified format across the entire humanities, transcending differences in specialized fields.

The Open Science Mandate and International Collaboration

In forming a digital scholarly space, the requirements universally demanded of all scholarly knowledge, regardless of field, include:

• Collaboration with experts from different fields, such as Research Software Engineers (RSEs) and Digital Librarians, in building academic infrastructure.
• Visualization of the knowledge construction process.
• Open Access to research results.

These requirements align with the philosophy of Open Science, which is currently being powerfully promoted worldwide, including in Japan and the West. These demands cannot be met by the individual-level efforts traditionally attempted in the humanities; they require a response at the level of the academic community. Furthermore, given that the academic space is not confined to a single country but is open internationally, it is necessary to build a system that can coordinate with the international academic community beyond the framework of domestic societies and adjust the construction of the academic environment to the level of international standards.

SAT’s Achievements and Global Network

The SAT Project, which serves as the core of this research, has steadily produced results regarding this series of challenges. The construction of the SAT DB was completed in 2007 with a budget of approximately 600 million yen, primarily from the Grant-in-Aid for Publication of Scientific Research Results and donations from the Association for the Promotion of Buddhist Studies, involving over 200 research collaborators. Since then, through additional Grant-in-Aid funding (Scientific Research (A), (S), etc.), we have mobilized another 200+ researchers to re-proofread the database and realized technical collaborations with leading overseas research institutions.

The SAT DB is a rare digital research environment, both domestically and internationally, that collaborates with digital research resources around the world and complies with international standards at a high level. Specific international collaborations include:

• Electronic Dictionaries:
  • Digital Dictionary of Buddhism (Charles Muller)
  • JapanKnowledge: Dictionary of Buddhist Terms (Commercial Japanese dictionary)
• Article Databases:
  • INBUDS (Japanese Association of Indian and Buddhist Studies, Domestic)
  • SARDS (University of Halle, Germany, International)
• Digital Catalogs:
  • The Buddhist Canons Research Database (Columbia University)
• English Translations:
  • Paragraph-based parallel corpus constructed in compliance with TEI Guidelines, based on the English Tripitaka (published by the Bukkyo Dendo Kyokai / Society for the Promotion of Buddhism).
• Digital Buddhist Images:
  • Sharing of research resources via IIIF in collaboration with the Bibliothèque nationale de France (BnF), National Diet Library (NDL), National Institute of Japanese Literature (NIJL), University of Tokyo, Kyoto University, and others.

Commitment to International Standards (ISO/TC37)

Regarding our involvement in international technical standards for the humanities, we have already detailed our contributions to Unicode, TEI, and IIIF in Section 1 (Academic Background). In addition, recognizing the need for detailed distinctions in language codes for handling Japanese Buddhist texts, we began participating in ISO TC37 last year. In collaboration with researchers from the National Institute for Japanese Language and Linguistics (NINJAL), we are preparing proposals for the formulation of international standards for historical Japanese language codes.

A Large-Scale System for the Future

As described above, regarding the challenges of “collectivizing digital infrastructure construction,” “visualizing the knowledge construction process,” and “making the space for providing and enjoying results public,” SAT has achieved remarkable results backed by collaboration among humanities researchers at home and abroad. In this project, we are organizing a large-scale system that mobilizes researchers not only from Buddhist studies but also from various humanities fields.

In the structure presented in [Figure 1], senior researchers guiding graduate students and young researchers across Japan will participate together in the construction of the CRDIH (Collaborative Research Database for the Humanities), aiming to build the new Daizōkyō. The technology used will be constantly exposed to review by cutting-edge domestic and international researchers and continuously reflected in new developments. Furthermore, we intend to share the insights gained from this research activity with humanities researchers in diverse fields across Japan involved in Digital Humanities, ensuring that the results are directly returned to those fields through review activities. Supporting the humanities as a whole in the digital age requires a comprehensive revision of the academic infrastructure that encompasses individual research, and this must be an endeavor that Japan is uniquely positioned to fulfill. This research is capable of achieving exactly that.

3. Project Origins, Domestic and International Trends, and Positioning

The Principal Investigator and the research team have long spearheaded research activities in Buddhist studies, actively promoting the digitization of the academic environment while working closely with the global Digital Humanities (DH) community. Through this process, we have come to strongly recognize the importance of shifting the research infrastructure of the entire Japanese humanities to digital media in compliance with international standards, while simultaneously presenting a pioneering research model for East Asian and Japanese studies. The history of activities leading to this recognition is outlined below.

(1) Developing Infrastructure Based on Evolving Digital Technologies

As mentioned earlier, we have continuously developed research infrastructure that adapts to constantly changing digital technologies. To supplement this: we have been invited not only to international symposia on Digital Buddhist Studies held annually worldwide but also to international symposia on East Asian Humanities and Library Science. Through these opportunities, we have provided knowledge on the latest technologies, their applications, and future possibilities to humanities researchers both in Japan and overseas.

(2) Compliance with International Standards and Collaboration with ADHO

Regarding the relationship between international standards and domestic research trends, our efforts have been described in previous sections. Through these activities, we have built a close international network with researchers in Western countries to stay updated on the latest applications of digital technology in the humanities. Simultaneously, we have formed a feedback network to relay this global knowledge to domestic humanities scholars. To return this global trend to the Japanese humanities, the Principal Investigator and team members are deeply involved with various domestic research and educational institutions as visiting or concurrent faculty, supporting their adaptation to international trends.

(3) Advancing the Framework of International Buddhist Studies

In the framework of international Buddhist studies, data linkage has progressed remarkably, creating highly favorable conditions to serve as a model for the humanities as a whole compared to other fields. The research team has led the construction of systems and data formats for major international projects:

• Indo-Tibetan Lexical Resource (ITLR): A joint project with the University of Hamburg.
• Digital Hôbôgirin: A joint project with the Académie des Inscriptions et Belles-Lettres (France).

These resources are published as open data on the European SSH Open Marketplace. Currently, international projects sharing academic Buddhist texts—such as the Buddhist Digital Resource Center (BDRC) in the US (Tibetan texts), CBETA and DILA (Dharma Drum Institute of Liberal Arts) in Taiwan (Chinese texts), and SARIT (Search and Retrieval of Indic Texts) by the University of Vienna and the Austrian Academy of Sciences—are aiming for high-level collaboration to improve the accuracy of automatic translation and text reuse detection technologies. Consequently, the necessity of constructing research data based on international standards—an area championed by our team—is being increasingly recognized. Amidst this trend, SAT members, with their proven track record in formulating international standards, are frequently invited to lecture at symposia and seminars around the world. [Reference Info]

(4) Responding to Changes in Humanities Research due to Generative AI

In text research, where major transformations are expected due to the emergence of Generative AI represented by ChatGPT, concrete measures regarding future technological development and regulation are already being formulated.

• April 2023 (University of Vienna): At the symposium “Advanced Computational Methods for Studying Buddhist Texts,” the Principal Investigator proposed a future direction centered on the “proactive yet critical adoption” of Generative AI and the rethinking of humanities methodologies, which was shared with the community.
• April 2024 (National Taiwan University): An international symposium on AI and Buddhist studies was held, where the proposal by our team reached a consensus.
• October 2024 (UC Berkeley): At a symposium hosted by the AI Research Lab at UC Berkeley, we presented and demonstrated a newly developed Buddhist studies research support system using RAG (Retrieval-Augmented Generation), showing the path for utilizing Generative AI in academic research.

In this research project, we will pursue a full-scale elucidation of this field by welcoming Sebastian Nehrdich, who is conducting cutting-edge AI research on Buddhist texts at the UC Berkeley AI Research Lab, as a research collaborator.

4. Scope and Goals of the Research

Core Goal: The “Reiwa Daizōkyō” and CRDIH

This research aims to collaboratively revise and re-edit the entire 85-volume “Taishō Canon” (comprising over 100 million Chinese characters) on the web, marking the beginning of the compilation of the “Reiwa Daizōkyō” (The Daizōkyō of the Reiwa Era). Simultaneously, we will construct the “Collaborative Research Database for the Humanities (CRDIH),” a new Buddhist Studies Commons that supports this endeavor. By treating this entire research process as a subject of study itself, we aim to provide the humanities community with insights into how researchers, information technology, and research workflows interact dynamically to construct humanities knowledge.

Technical Innovation: The SATed System

Prior to this project, the Principal Investigator and the team developed SATed, an AI-driven OCR collation system that boasts high accuracy in processing images of Buddhist manuscripts and printed books. SATed has now reached the practical application stage.

• Function: SATed automatically compares high-precision text data generated by OCR software with the text data of the SAT DB (which has been manually refined by the team) character by character.
• Output: It highlights discrepancies, presenting not only the differences between OCR data and SAT DB data but also the variant readings (collation data) between the SAT DB and various manuscripts/woodblock editions.

Transforming the Research Workflow

In traditional textual collation, researchers had to visually verify the layout character by character. Even with the spread of digital images, the low accuracy of OCR for East Asian classical texts made this an extremely labor-intensive task, often paralyzing researchers before they could even begin analysis. SATed dramatically reduces this burden by presenting highly accurate comparison information. This allows researchers to view discrepancies at a glance and focus on their primary high-level task: interpreting these differences. The CRDIH provides an environment where the historical development of Chinese Buddhist texts, which spread from East Asia to Central Asia over 1,500 years, can be examined and judged efficiently.

[Scope of Materials] As shown in [Figure 2], the CRDIH will integrate and automatically compare (via SATed) the following resources with other diverse research materials:

• Approximately 150,000 to 190,000 images of woodblock prints from the various editions used as the basis for the Taishō Canon.
• Images of Dunhuang manuscripts.
• Various manuscripts containing iconography.

Transparency, Evaluation, and the “Preprint” Model

We will record the entire academic process within the CRDIH to clarify the locus of responsibility for revisions, evaluate individual researchers, and visualize the knowledge construction process.

• Process: Researchers visually verify and report on text collation work.
• Quality Assurance: To confirm and evaluate academic quality, we will establish the CRDIH Editorial Board, composed of domestic and international experts. This board plays a core role in ensuring quality.
• Open Sharing (Preprints): Crucially, regardless of whether data is officially authorized by the Editorial Board, the collated data will be released and shared under the responsibility of the collator. This aligns with the philosophy of “Preprints,” which is becoming increasingly important in modern academic information circulation.

Formulating Guidelines for the Humanities

The insights gained from establishing and operating this system will be vital for future web-based scholarly editing across disciplines. Therefore, we will record and analyze these findings to create “Guidelines for the Humanities,” which will be published as Open Access data. While the Japan Consortium for Open Access Repository (JPCOAR) provides excellent guidelines for Research Data Management (RDM), they alone are insufficient to address the unique circumstances of the humanities. This research aims to fill the gap between general-purpose guidelines and the specific needs of the humanities.

Outreach and Sustainability

To return these results to the Japanese humanities community, the “Humanities and Informatics Development Committee” will meet online every two months and hold an in-person research meeting once a year. These meetings will discuss the project’s status and feedback findings to various humanities fields. Furthermore, the collation of the Daizōkyō DB must continue into the future. The significant strength of digital media lies in the fact that as research progresses, the source materials themselves remain exposed to criticism and are continuously revised.