Sharaf al-Din's positive-root analysis for cubics

Before trying to calculate an answer, ask whether the problem can have the positive answer it requests. This distinction is central to the contribution attributed to Sharaf al-Din al-Tusi in the biographies' account of cubic equations. Some classes admit positive solutions under their conventions; others depend on coefficient values. One failed numerical attempt does not establish impossibility, and calling something an equation does not guarantee the kind of solution sought ¹.

This record concerns that conditional analysis, not all algebra or another count of Khayyam's geometric constructions. A maximum supplies a solution threshold in a clear example. Our symbolic demonstrations are independent modern explanations, not manuscript transcriptions. Historians' differing interpretations of the relation to derivatives remain attributed, not converted into a claim that complete modern differential calculus already existed ^{2 3}.

This is a method for analyzing equation classes, not an instrument or an isolated numerical table. The accounts describe twelve lower-degree types, eight cubic types with positive solutions, and five cubic types whose positive solutions can fail to exist. Positive coefficients and positive roots belong to the conventions. The classification is not a purported inventory of every complex solution in modern algebra ^{1 4}.

MacTutor's example is x³+a=bx with a,b positive. Rearranged in modern notation, a=bx−x³ asks whether the expression on the right can attain the requested value for positive x. The possibility question can therefore precede approximation. These printed letters and symbols are modern presentation: the biography explicitly says Sharaf expressed his mathematics in words ¹.

Cubic problems already had a geometric history. Ozel places Sharaf after Khayyam's treatment through conic sections. That does not mean they did exactly the same thing or invented all curves. The article distinguishes numerical determination and investigation of solutions while retaining geometrical proof. A separate record is justified by the particular existence-and-threshold question, not by retelling an earlier intersection construction ².

MacTutor also identifies earlier Arabic-language techniques for approximating roots of numbers before discussing their broader equation applications. Such antecedents do not create a single inventor of every approximation method or prove a direct chain to Horner. They locate a field in which a specific extension could occur. Scope and attribution still need the reference's stated limits ⁵.

The Treatise on Equations is described as surviving in an anonymous abridgement. According to the accounts, numerical computation tables were removed. That changes what can be reconstructed: classification and argument can survive while parts of computational implementation disappear. We do not fill those gaps with generated numbers and label them Sharaf's original tables ^{4 6}.

Ozel records Rashed's1986 Paris edition and a1998 Arabic translation by Fares. These are bibliographical reports from a read source, not editions independently read here. Ozel also warns that a1209 lifetime claim resulted from a misread erased manuscript date. We do not make that an exact discovery or composition year. The record uses a twelfth-century horizon, not a documented invention day ⁷.

Sharaf al-Din al-Muzaffar al-Tusi is the teacher associated with cities including Damascus and Mosul, whose life is sparsely documented. He is not Nasir al-Din al-Tusi or his own pupil Kamal al-Din ibn Yunus. A familiar surname does not establish identity. MacTutor connects the algebraic work with later Baghdad teaching, not the discovery site of every proposition ^{8 9}.

MacTutor explicitly identifies the individual as an Islamic mathematician. That supplies the Muslim basis rather than his name or patron. Sunni remains separately the museum's declared default frame; no personal school was established by these readings. The anonymous abridger and modern interpreters also have roles: transmitting and explaining a method does not automatically make each transmitter its mathematical author ^{10 4}.

MacTutor explains that bx−x³ reaches its maximum at t=√(b/3), with maximum2(b/3)^(3/2). A positive solution in the example therefore requires a not to exceed that bound. Since a is positive, a root must also lie between0 and√b. “No solution” in this setting means no requested positive root, not an assertion that all negative or complex roots disappear ¹.

Here is an independent modern proof without differentiation. Put b=3t². Then2t³−(bx−x³)=(x−t)²(x+2t). For x nonnegative and t positive, the right side is nonnegative, vanishing only at x=t. This proves a global bound on that domain, not merely the largest value among sampled guesses. It is our algebraic demonstration, not a claimed reconstruction of the manuscript's proof sequence.

Ozel describes reception among later authors of algebra and arithmetic, naming works that refer to Sharaf's treatment. This supports a bounded claim of literary and mathematical reception, not a census of students or schools everywhere. An author's use of the book does not establish that every algorithm or every subsequently omitted table was adopted ¹¹.

Modern editions and studies brought the work into further historical debate, including differing derivative interpretations. The original studies were not independently inspected here; their positions are reported through the two biographies. Reading a bibliography is not reading all the works it lists, and the later names Newton or Horner in an analogy are not documents of a continuous transmission route ^{3 4}.

The accounts describe equations, geometrical reasoning and numerical approximation working together. The comparison with Khayyam makes the added question visible: under which conditions can a class have the positive solution sought? Separating existence from computation is a way to explain this particular contribution, not a verdict that every earlier mathematician ignored whether a construction could work ².

In our modern example choose b=3. Then3x−x³ has maximum2 at x=1. Choosing a=1 lies below the threshold, a=2 on it, and a=3 above it. These are teaching choices, not historical table entries. Keeping positive x explicit prevents the mistaken idea that a missing positive root means no real number of any sign can satisfy the equation.

For a=1, our example has two distinct positive roots: one between0 and1, another between1 and√3. This modern conclusion follows from continuity and the rise then fall of3x−x³ on the relevant positive domain. It does not claim Sharaf used a later formal intermediate-value theorem. Bracketing and narrowing a root here are explanatory computations, not recovered rows of an omitted table.

At a=2, x³−3x+2 factors as(x−1)²(x+2). There is one distinct positive root,1, with multiplicity two, and a negative root−2 outside the positive-root question. At a=3 there is no positive root because the maximum2 cannot reach3, although a negative real root remains. Distinguishing distinct roots from multiplicity prevents an apparent contradiction in counting solutions.

This is a contribution to knowledge, not a pump lifting water or a clock sounding in a market. The sources describe deeper equation analysis and reception among mathematicians, not an economic ledger of trade transformed by a particular bound. We therefore do not turn a proof into an unsupported crop-yield or navigation claim. The directly explainable benefit is organizing a question before undertaking a futile search ¹².

For instance, requesting a=3 with b=3 will not be rescued by trying positive guesses forever. The obstacle is the upper bound, not a poor approximation table. This is a modern educational illustration, not testimony from a twelfth-century user's diary. Likewise, calculating more decimal places for a possible root cannot fix having chosen the wrong equation class or the wrong requested domain.

Ozel presents the work as important after Khayyam and compares its questions with existence, maxima and approximation in modern language. Comparison must remain comparison; later mathematical names are not terms found in the author's own wording here. Differing interpretations show why history requires more than matching an old expression to a modern classroom rule ^{2 3}.

The supported contribution is narrower than inventing all algebra or single-handedly anticipating every computer. It is systematic analysis within an equation tradition, preserved through an abridgement missing tables and reflected in later mathematical reception. Those limits preserve its place among earlier practices and later interpretations without erasing any of them ¹³.

Sunni museum frame: we attribute Sharaf's method through Ozel and MacTutor, separating it from Khayyam's earlier constructions and modern symbolic notation. No universal priority is claimed for all maximum problems or numerical methods. Nor is this method the same contribution as his linear astrolabe merely because the maker is shared ^{14 8}.

Academic interpretation: MacTutor reports differing explanations of the maximum procedure, including a derivative analogy and an alternative account. We present the disagreement through that source, not as though Fares's and Berggren's original papers were independently read. The museum neither awards a blanket “first calculus inventor” verdict nor dismisses sophisticated analysis because it used a different language ³.

Readers can independently check the bound, plot examples and compare approximation with exact factorization. These are modern activities in this presentation. An equation can be rearranged in different forms, and classifications restricted to positive coefficients need not have the same type count as classifications allowing signed coefficients. Domain conventions matter when comparing old work with modern algebra ¹.

The displayed image is a different modern polynomial, with roots−4,−1,2 and equation(x+4)(x+1)(x−2)/4. It is not3x−x³ and not Sharaf's drawing. It supplies cubic-curve context while stating the mismatch. Its metadata records N.Mori and the public-domain release; the page does not present every cubic picture as medieval, anonymous or a graph of the worked example ¹⁵.

The textual witness described by the biographies is an abridgement without the numerical tables. Ozel discusses comparison with another work to assess the abridgement. We know that through the reference, not a collation performed here. Reports of1986 and1998 editions do not license invented page citations to unread volumes or the attribution of modern calculations to the uninspected manuscript ⁴.

A mathematical method does not need a surviving industrial artifact to count as knowledge. The image and chain are explanatory media; the historical evidence concerns the text and reception. A modern plot presented as an original page would change the evidence's meaning. That is why the graph's authorship and different equation remain visible, alongside the source-reading limits ^{16 4}.

Algebra's history can tempt readers into a race toward modern labels: Newton, Horner, derivatives and algebraic geometry. The biographies themselves use such comparisons. Our responsibility is not to convert an analogy into ownership of a complete later theory. A mathematical relation can agree without the authors sharing purposes, language, teaching context or a documented path of textual transmission ¹⁷.

Knowledge also passed through pupils, abridgers, commentators and editors. Preserving their roles and what remains unknown is part of the history, not clutter. The declared frame likewise distinguishes individual Muslim identification from mathematical proof: proving a maximum establishes no religious school, and possessing a modern translation would not automatically resolve every disputed life date ^{18 19}.

Start with a smaller question: for positive x, can3x−x³ equal3? Our modern identity shows that it cannot exceed2. You do not need a thousand guesses to answer. Asking for1 allows two crossings; asking for2 reaches the top at one distinct positive value. These are paper numbers, not a measurement experiment or a claim that a historical table contained our choices.

The larger lesson is to check whether a target is possible before hunting for its location. Positive and negative answers have different relationships to the question's conditions, not different moral worth. Do not copy the accompanying graph as our example: its stated polynomial is different. Distinguishing the picture, the historical claim and the modern explanation makes a more careful reader, not merely a collector of invention names ¹⁶.

• ¹⁷ O'Connor and Robertson,complete biography,1999.
• ²⁰ Ahmet Ozel,biography,analysis and works,2012.
• ¹⁵ Different modern public-domain graph;original equation editions and interpretation papers not independently read.