The Evolution of Islamic Law1 and Text Recommendation

This is a "Natural Language Processing" project that aims to look at how the four main madhāhib² have both individually and collectively changed over time — from their inception by the "founders" themselves up until the current-day scholars who still produce works of Fiqh (Islamic Law).

This project will use machine learning and the law books of the four schools (found on http://shamela.ws/ in their respective sections).

Problem Statement

I hope to be able to answer the following questions:

1. Which words and phrases are the most specific to each school? (i.e., do Ḥanafī³ works often cite “qiyās”⁴ far more than Ḥanbalī⁵ works do?)
2. How do those school-specific words and phrases change from century to century? (i.e., do the earlier works predominantly quote Qurʾān⁶ and Aḥādīth⁷, while later works simply reference back to the opinions of previous scholars?)
3. Can I provide valuable book recommendations that would leave the reader satisfied with the suggested texts?
4. [To be addressed in future iterations] Can these keywords and phrases be used to accurately predict with which madhhab an author affiliates? (i.e., do the ḥadīth commentaries of Mullā ʿAlī al-Qārī [rḥA] and Imām al-Nawawī [rḥA] clearly mark them as “Ḥanafī” and “Shāfiʿī”⁸ respectively?)
5. [To be addressed in future iterations] Can I provide a summary of the texts in a manner that is both inclusive and exclusive, but still staying true to the original intent of the author?

Note:

1. The years and centuries used in this project refer to the Islamic calender, i.e., the Hijrī Lunar calender. For a quick conversion, it will be 578 years behind the current Gregorian calender.
2. I understand that limiting myself to only those books found within Shamela is not the best representation of the schools as a whole, but this is where there are data available for such a project.
3. I also understand that there are many books which have commentaries (i.e., for al-Marghinānī’s al-Hidāyah, there are at least 3 massive commentaries which include the entire text of al-Hidāyah: al-Bināyah, al-ʿInāyah, and Fatḥ al-Qadīr) which might skew the results. In the process, I will search for the optimal statistical and computational method in which to address this issue.

Project Timeline

Data

• 4 Categories
• 251 Books
  • Ḥanafī: 48
  • Mālikī: 66
  • Shāfiʿī: 60
  • Ḥanbalī: 77
• Nearly 500K Pages ~5GBs

Footnotes

_{^{1. Fiqh is the term used for implemented Islamic Law.}}

_{^{2. Plural of madhhab. Refers to one of the 4 legal schools of Islamic Law.}}

_{^{3. One of the four schools. The first to be codified and well documented (which occurred in the late 2nd century).}}

_{^{4. Legal Analogy}}

_{^{5. One of the four schools.}}

_{^{6. Muslim's religious book, believed to be the words of God.}}

_{^{7. References to the statements and actions of Prophet Muḥammad ﷺ (d. 633 CE).}}

_{^{8. One of the four schools.}}