Method

What does this page explain?

Here we write how we did the work, what we deliberately did not do, and where we stand. Every sentence on the pages that follow comes out of this method; read without knowing it, they may sound as if they claim something they do not.

Let us say it at the outset: this is not a summary of readings, it is our own framework of reading. We do not understand Dost; and we do not give up trying to understand him.

How did the question change?

At first we asked the question wrongly. We were asking: "Is there anything in physics experiments that matches the mechanism Dost describes?" This question pushed us toward matching mechanisms, that is, toward fitting one thing to another. On three topics (re-creation at every moment, self-disclosure, the isthmus and imagination) we looked this way; we found no counterpart at the level of experiment or theory. We did not have those answers deleted, but we set the question itself aside.

The new question was posed by Murat: we are reading Dost; we note the places our reason cannot take in. Quantum physics also has facts that reason cannot take in. Is the way they fail to fit reason the same in both? When we place them side by side, does the question "could this be the same thing?" arise?

This question has a footing: neither side says "we understand." In Dost this poem stands (together with its second line):

«We were bewildered, and so was He / Whoever is bewildered has not walked.» (c15k169, p. 416)

We do not use this poem as a basis for saying that both sides say "we do not understand"; attributing bewilderment to al-Haqq would not be our claim to make. The footing for the common ground is the words on the physics side and the stance of these pages themselves: we see physicists of the founding generation speaking from the same place; Feynman wrote that one can safely say that nobody understands quantum mechanics [Feynman 1965]. The field's own survey also shows that its foundational questions are still hotly debated [Schlosshauer 2013]. This does not make the two sides equal. It only shows that, on the physics side, we begin from the same place.

The criterion of the new question is form. That is, we look at this: how does something fail to fit reason? Are two opposites true together? Does the way of looking determine what is seen? Does the same thing not come a second time? We do not look for a match of mechanism; not finding one is not a ground for rejection either.

How did we read Dost?

We read the Futuhat, all 226 sections, from beginning to end, from the site's Turkish section summaries. In total it is a text of about 2.19 million characters. We divided the reading among reader agents working in parallel; a reading record was kept for each section.

While reading we looked for one thing only: passages in Dost that seem not to be solvable by reason alone. We called these "enigmatic items"; this is our working label, not a quality we attach to Dost's words. Sentences that appear contradictory, things told through images, places where "the people of unveiling know," places left in bewilderment were included. Only rulings of jurisprudence or theology, history and anecdote, and the manner of worship were left out.

One decision was important: we did not filter by keyword or quotation marks. An enigmatic passage often stands unquoted, inside a plain sentence; a hunt for words would have missed it.

The result is 2,788 enigmatic items across 226 sections. To each of them we gave a difficulty score: 1, "easily likened"; 3, "we struggle greatly to find a likeness." 1,154 items received difficulty 3. This is the main pool we lean on in the pages.

Then we grouped the items by their forms. The first reading yielded fourteen forms, all of which are seen across all 18 volumes; in the second reading Multiplicity and Reciprocal transformation were added. These forms appear on the pages under their short names (their expansions are in the introduction page): Both-and, Non-repetition, Continuity, Relation, Viewpoint, Wholeness, Cause, Containment, Time, Knowledge, Reason, Indistinguishability, Inside-outside, Bewilderment, Multiplicity, Reciprocal transformation.

The second reading was done as follows: all 226 sections were read again, with the Multiplicity and Reciprocal transformation forms in mind, divided among nineteen reader agents. The quotations were automatically compared with the source text; the two items that did not match were removed. In the second reading we also flagged forty-four new patterns; these have not yet been read side by side with physics.

Wherever we quote Dost, the section number and, where possible, the page number are given. The quotations were taken from the reading record, automatically compared with the source text. The 16 quotations that did not match in the comparison we stopped counting as "quotation" and marked as "summary." We did not count the site's own "That is…" explanations as Dost's words.

How did we read physics?

We tied every physics sentence to a source. The number of sources is 97 (for the order of the sources and the range of view, see the limits below). All of them are sources whose bibliographic details were checked one by one through Crossref, arXiv, or the address. We used no citation that could not be verified; we invented no citation.

Our order of sources: peer-reviewed articles and reviews, then arXiv preprints, then encyclopedia entries (such as the Stanford Encyclopedia of Philosophy). We did not use popular accounts as sources. The sources are physics literature and encyclopedia entries; the literature of Ibn Arabi specialists (the commentators) and the critical literature on physics–Sufism comparisons were not used.

In recounting, we kept three things apart.

• Experiment: a measured, repeated result.
• Theory: the mathematical content of the theory.
• Interpretation: giving meaning to this result. Here there is no consensus among physicists; interpretations such as Copenhagen, many worlds, and Bohm say different things about the same experiment. Whichever interpretation a similarity depends on, we wrote it as dependent on that interpretation.

We also marked untested theories separately as "speculative."

How did we place them side by side?

Every form page has the same order: first Dost, then physics, and last the side-by-side look. In the side-by-side look we wrote four things, in order and in plain sentences: what is common, where they part, where someone who sees might notice this, and what question we are left with.

We applied three tests to every closeness.

1. Specificity: "What other oddities does this form also fit?" Every odd thing resembles every odd thing; this is our greatest trap. If a form was not specific enough, we lowered the closeness. A crowded form (for example, opposites being together) was not counted as specific, because it appears in so many places.
2. Where they part: Written openly in every pair. In two places the parting was so great that we counted it as a finding in its own right.
3. Counter-reading: For each closeness we thought through and wrote what a skeptical physicist and a commentator on Dost might say.

What do the words of closeness mean?

We use five phrases. None of them declares the two sides identical.

• They stand close. The manner of telling on the two sides looks toward each other, and the form is specific enough. A question can be asked here. We do not have the answer.
• They greet each other from a distance. In both places something says "it is not as it appears," but when we go into the detail, they seem to be speaking of separate subjects. We count it as a greeting, not as a closeness.
• They walk separate roads. The two sides go in opposite directions. We write this too as a finding.
• We did not look at this door. We did not look for a counterpart, or we looked and did not find one. "We did not find" does not mean "there is none"; someone who sees may think differently.
• Tempting, but we did not do this. There is only a similarity of words. We left it aside deliberately, so as not to take a similar-sounding word for a similar thing.

What did we not do?

• We did not end with a claim of equality. Every page closes with a question.
• We did not say that physics confirms Dost or that Dost describes physics; in neither direction.
• We did not add our own words to Dost's words. The quotations stand as they are.
• We did not fit physics to Dost. Where there is a divergence, we wrote it.
• We did not look for mechanism; we looked only at form.
• We do not claim to have found a criterion that separates a deep closeness from a superficial one.

Limits

It is necessary to say these openly.

We read summaries. What we read is not the Futuhat itself, but the site's Turkish section summaries. Dost's words that fall outside the summary are not visible in this reading. That a sentence does not fit reason may also come from how the summary conveys it.

Comparison with the Arabic original is limited. We compared only the first thirty quotations with the Arabic (twenty-three match, three partly, four could not be found); the rest remains at the level of translation. Every quotation on the pages comes from within a translation, and that in turn from within a summary. The Arabic text we compared against is the digitized edition in OpenITI (Dār Iḥyāʾ al-Turāth al-ʿArabī, Beirut 1998; OpenITI licence CC BY-NC-SA 4.0): a single edition, it may contain digitization errors, and the text is not published on this site.

The translation check is automatic. We also ran a check against Ekrem Demirli's 18-volume translation. We took the 1,087 difficulty-3 items carrying quoted fragments; we searched each quoted fragment in the relevant volume of the translation, ignoring differences in letters, accents, and punctuation, with windows of 24 characters. Result: 937 items were found, 71 were found partly, 79 could not be found. Of the unfound ones, we examined 34 of the quotation type one by one by hand. No real false quotation emerged. The reasons fell into three: spelling corruptions coming from scanning, the summary shortening and adapting the sentence, and the rendering of a Qur'anic verse being different.

Two things to note here. First, the check is computer matching; it is not a reading for meaning. Second, this is not an Arabic check: whether the translation conforms to the Arabic is a separate task, and we did not look at it. We have not put the text of the translation itself on the site or in the repository; only the result numbers stand.

Some of the physics summaries. For some sources, we wrote the summary not from the full text of the article but from general knowledge. We marked these as "uncertain" on the relevant pages.

The forms are not specific to Dost. Most of the forms we chose exist in other traditions too (Nagarjuna, Eckhart, Heraclitus; in kalam, such as no repetition, the means (vesile), nothing better being possible). We did the specificity test here only on the physics side ("what other oddities does this form fit?"); we did not try the same matching with a non-Dost text.

Multiple comparisons. Among nearly a hundred physics sources, we chose the best matches. The more things we look at, the easier it is for something to resemble by chance; this is a limit.

Kalam context. We did not separately read concepts of kalam origin, such as no repetition, means (vesile), and time by relation, within their kalam context.

The share of interpretation is ours. Which item we counted as enigmatic, what difficulty we gave it, which form we put it in are our decisions. Two other readers might produce another list.

What could come next?

Going back to the Arabic original and checking the hundred or so most enigmatic items could be a next step. And there is your own reading: if someone who sees notices what we could not, these pages were written for exactly that. We look forward to your reading and telling us.

The blind-matching trial

When we say "these two things are similar" on every form page, we had to ask ourselves whether we were really looking at similarity of subject (time with time, knowledge with knowledge) or similarity of structure. To test this, we ran a small trial.

We gave the description of each form on the page to an agent and asked it to strip the subject away and turn it into a plain structure: like "X is both there and not there," without words like time, knowledge, al-Haqq, servant. Then we gave a second agent only this structure and short descriptions of a hundred and one physics sources, and said "choose the six sources that best fit the structure." We looked at how much this overlapped with the sources we had tied to that form on our pages.

The result: across sixteen forms, ninety-six choices were made in total, and twenty-nine overlapped with the source on our page (thirty percent). By random selection we would expect six percent. That is, the structure alone, even independent of subject, found our matching far above chance. But the distribution is not even: in the forms Time (six of six choices), Relation, Containment, and Knowledge there are four hits each; in the forms Both-and, Continuity, Inside-outside, Indistinguishability, and Bewilderment there are no hits at all. In those five forms, the matching on our pages may come more from subject or from our own interpretation than from structure; we do not know this.

This trial is not solid evidence: it ran once, both agents were AI models, the structure descriptions could not be fully stripped of subject (hints such as measure and direction remained in the description of the Time form), and our source sets had themselves been chosen by looking at the subject. Still, it showed two things: some matchings can be read from structure, some cannot. The ones that cannot need to be stated more cautiously on the pages.

Independent review

We turned all the matchings on the pages one by one into records (a hundred and seven records; each with a mandatory answer to "for what reasons might this similarity be wrong?") and had these records reviewed by a separate reader who had not seen the pages. The records and the method stand on the research side; they do not enter the site.

The result of the review: fifty-two of the records were kept, thirty-six were told "should be lowered by one grade," nineteen "should be rejected." We divided the similarities into five classes (A: counterpart in the same technical sense, B: structural, C: conceptual, D: figurative, E: superficial or rejected). Today no record counts as A or B. The strongest are in class C (eleven records), four records are divergences, the rest are figurative, superficial, or doors we did not look at. We tried to keep this consistent with the phrases on the pages: no matching remains of which we could say "they stand close"; at most, the pages say "they greet each other from a distance."

This is not a failure; it is the honest state of the work: the ways the two fields fail to fit reason can be placed side by side, and some questions are common, but we have not found a matching solid enough to call "the same structure." The questions on the pages therefore remain questions.