Jalali calendar reform

Imagine a calendar whose month numbers stay orderly while its beginning slowly moves through the seasons. The Jalali reform sought to anchor New Year to the spring equinox. Supported by Malik Shah, it took its name from his title Jalal al-Dawla; Maliki and Malikshahi were alternative names. It did not invent the solar year, and its history is not one person's discovery of time ².

The project is commonly associated with Umar Khayyam, whose participation MacTutor reports. Johannes Thomann, however, asks whether named observers can automatically be identified as the calendar's designers. We retain that disagreement, distinguish the epoch used for counting from an unknown invention day, and separate later leap rules from an original programme we cannot reconstruct completely ^{3 4}.

A solar-calendar reform connects the year's beginning to a seasonal astronomical event and reconciles a whole-number day count with the solar year by adding days in some years. Iranica describes New Year in relation to the sun's apparent entry into Aries before noon. Here the zodiacal term specifies an astronomical reference, not a claim that a constellation governs someone's life ⁵.

A solar year does not make every month begin at the sun's entry into another zodiacal sign. The source distinguishes astronomically anchored years/seasons from historical thirty-day months, with additional days completing the year. Nor should the modern arrangement of six thirty-one-day months be projected back onto the eleventh-century reform ⁶.

Persian calendrical history long preceded the Seljuks and Islam. Thomann explains a system of twelve thirty-day months plus five additional days:365 days, roughly a quarter-day shorter than the seasonal year. An accumulating small difference moves month dates through the seasons, even while names and numbers remain perfectly orderly in an account book ⁷.

That movement mattered socially. Reform narratives complain about tax payments falling before crops supplied the means to pay. Earlier Abbasid and other attempts are reported, but Thomann tests such narratives against recorded dates rather than treating every reform story as permanent implementation. An old need and a proposed remedy do not establish how successfully people adopted it ⁸.

Thomann describes Malik Shah's summons to astronomers from Khurasan and the observing work surrounding the project. Yet the Nawrūznāma, attributed to Khayyam but probably not his work in the historian's assessment, says intercalation was not completed. A historical account discussed alongside it distinguishes reform from observatory construction and the end of observations after the sultan's death ⁹.

We therefore invent neither a founding meeting with recoverable dialogue nor an original calendar notebook completed on a known day. The witnesses Thomann discusses contain1076 and1079 epochs, with a later tradition preferring the latter. Iranica interprets468 differently, as a fiscal-year number rather than the same Hijri year. The disagreement over how to read a number belongs in the history, not in a silently harmonized chronology ^{4 5}.

Declared Sunni frame: This is the museum's frame in Khayyam's person record, not testimony about his personal school or an inference from a patron's name. Individual Muslim identification rests on Stanford's description of his philosophical writing; other attributed readings remain visible in the person record. Identity evidence cannot determine which part of calendar work he performed ¹⁰.

Reported participation and historical question: MacTutor says he contributed to the reform. Thomann instead notes anonymous reformers in the reform account and named astronomers in the observing account; Khayyam could have worked on observations or tables rather than the design itself. We qualify the recorded association, acknowledge the sultan's patronage, and do not turn an observer list into a verified inventor list ^{3 4}.

Begin by separating event from count. The equinox is an astronomical event; the day is a calendrical unit, and a written calendar year has a whole number of days. Connecting the beginning to the event requires a rule for choosing the day and accounting for common and leap years. Calling something “accurate” without specifying the epoch, months, added days and version tells us too little ⁵.

In our modern paper model, not assigned to a particular manuscript, twelve30-day months make360 days; five extra days make365. Eight added days over33 years give365×33+8=12053 days, averaging365+8/33. This illustrates addition and distribution, not a securely recovered Khayyam algorithm.

Compare a model adding one day every four years: its average365.25 yields12053.25 days over33 years, a quarter-day more than the other model. Arithmetic difference alone does not establish perpetual agreement with the sky. Iranica presents differing later witnesses and rules, including four/five-year intervals, not a warrant to collapse all of them into one fixed original cycle ⁵.

Indeed, its listed five-year-interval years include a transition from163 to192: a gap of29, not33. Readers can check the subtraction to see why “33-year cycles occur” is not equivalent to “every cycle is always33.” Historical conversion needs a dated text and a specified rule before it becomes an algorithm ⁵.

A description in astronomical tables does not prove use in every administrative office. Thomann looks for inscriptions, coins and documents; his Jalali examples occur particularly in astronomical/astrological materials, challenging the picture of comprehensive administrative replacement. This is a scholar's conclusion from his evidence, not proof that every unobserved use was impossible ¹¹.

In fragments dated1182/1183, he infers an equinox-anchored year while older Persian dates remain inside the table. An Ottoman example from1450 has a separate Jalali column rather than one exclusive date system. Here circulation concerns neighboring ways to count time: a calendar can travel within a specialized practice without erasing others ¹².

The reform inherited month names, eras and calculation practices rather than starting from nothing. Reconciling solar motion with numbered days and attempts to realign seasonal festivals preceded it. Thomann reconstructs that setting through studies and historical accounts; this article does not claim independent reading of every Persian original or inscription in his references ¹³.

Observations and tables formed the working environment, but their presence does not establish a technical chain from every astronomical instrument to this calendar. We add no unsupported causal link from Khayyam's cubic constructions to the reform, and do not make a calendar reform the invention of observatories or observation itself ⁴.

Later astronomical tables discussed its epoch, intercalation and conversion with varying formulae. Iranica describes stages of this treatment and modern changes to the chosen era, month names and month lengths. Historical continuity does not mean every rule stayed identical across centuries ⁶.

It associates1925 with changed Solar Hijri month lengths: six months of31 days, five of30, and a final29 or30. We do not transplant this structure into1079, or present the article's historical formulae as current legal conversion software. Calendar, era and particular version must be distinguished before two dates can be compared responsibly ¹⁴.

A complaint about taxes and crops supplies an understandable motive, not a Jalali tax receipt. Thomann separates motives reported in literature from implementation demonstrated in documents. Respecting that difference avoids imagining farmers whose lives were transformed by an order for which we lack that direct kind of evidence ¹⁵.

Multiple dates in astronomical documents could support comparison and recognition of the same day. The historian also examines astrological functions in scheduling personal choices. We attribute that interpretation, without converting the document users' beliefs into predictive knowledge endorsed by the museum or advice to follow such practices ¹¹.

The project's significance lies in equinox anchoring and its long afterlife in astronomical writing, not a slogan about the “most accurate calendar ever invented.” Its history becomes richer when patron, observer, scribe and user are considered separately rather than compressed into one famous signature ¹.

The programme/implementation distinction also teaches us how to read innovation. A book may describe an orderly system while an everyday document retains an older convention. Finding both is not a reason to erase one; their coexistence opens a question about actual practices of use ¹⁶.

Reported attribution: MacTutor gives Khayyam a familiar association with the project, but the reform is collective and Malik Shah the patron. This supports neither sole invention of all solar calendars nor ownership of every later leap rule ³.

Historical question: Thomann questions how observing narratives establish Khayyam's role. We do not replace popular certainty with opposite certainty excluding him. Sources also agree on1079 while differing over the earlier epoch and Hijri day:9 Ramadan in Iranica versus10 in Thomann. The record uses471 and Julian1079 as the preferred later epoch, not a verified invention day ^{4 5}.

The modern Solar Hijri calendar belongs to this heritage but is not a frozen copy of the medieval reform. Iranica discusses choosing the Hijra era and modern legislative changes. A solar year does not imply abolishing lunar Hijri dating or transferring its religious observances into a classroom arithmetic rule ¹⁴.

Tables still invite verification: which era, an old or Jalali month, and which leap-year rule? A matching month name is insufficient, and many decimal places do not guarantee everlasting accuracy. The Berlin example shows why headings and neighboring columns must be read before a number is converted into a date ¹².

The displayed image is folio1r of Berlin's astronomical-table manuscript Ms.or.quart.211, dated1804 by the institution. We inspected its original image: red ruling, entries in black/red/green, and library stamps. The institution records six Arabic leaves and Public Domain Mark1.0; this is not an original leaf in Khayyam's hand ¹⁷.

Thomann identifies seven parallel calendars in this example, explaining that the original first leaf is missing and the date is established astronomically. We follow his column interpretation, not a claim to decipher the complete manuscript independently. Conversely, an astrolabe he discusses from an older description has no known present location; we do not claim it is on display in a visitable gallery ¹⁸.

Nowruz, royal naming and multilingual/multicalendar writing show that a date can express affiliation as well as a number. Thomann reads colored multiple dates in Iskandar Sultan's manuscript as chronological eloquence, not merely calculation. This is his interpretation of a particular object, not a judgment about every calendar user ¹².

Nor should Khayyam's poetic image authenticate a technical account. Stanford's distinction between philosophical and poetic voices helps resist that compression. We infer no personal school from a disputed quatrain and reproduce no poetry as a substitute for documentation of the reform ¹⁰.

Write365 spaces for a paper year and distribute eight additional days across33 year cards. Work out the total and average, then move the additions: an identical average does not uniquely determine which years are leap years. This is a modern arithmetic activity, not a recovered ancient table or a rule for dating religious observances.

Look at the Berlin image: why might one page contain several calendars? Try writing today's date in two systems your family knows, naming each system. No sun-gazing or observing apparatus is needed; the aim is to understand how different counting conventions can share one document ¹⁸.

Thomann2021, actual accepted-version main sections1–6, not all appendices; Reza Abdollahy's Iranica Jalali/Solar Hijri subsections; MacTutor for reported participation and Stanford for identity context; institutional Berlin metadata, folio1r and rights. No source manuscript or protected text is republished, and not every listed original was independently read ^{19 6 3 10 20}.