Corrections Log

Errors found in this encyclopedia are corrected in the article and recorded here. Each entry states what was wrong, what replaced it, and why.

This log exists because a reference work that hides its errors gives readers no way to judge its reliability. The entries below are not comfortable reading. They are the reason the current text can be relied on.

August 2026 — Accuracy Review

A review of the encyclopedia's Puerto Rican coffee articles examined every claim about the island's historical customers, its export figures, and the brands that carry its heritage. Sixteen articles were corrected. The principal findings follow.

The sixteen are those carrying claims about papal, royal, or brand heritage. A further two articles had links updated without changes to their claims, and separate corrections to other articles are recorded in earlier and later entries.

1. The Pope Leo XIII endorsement — removed

What was published. Several articles stated that Puerto Rican coffee, particularly from Yauco, was endorsed by Pope Leo XIII, who reigned from 1878 to 1903. One article stated: "Historical records confirm that Pope Leo XIII drank Yauco coffee regularly."

What we found. No historical source for this attribution could be located — no primary record, no scholarly work, no archival reference. The claim appears widely in retail and tourism material but traces to no evidence. The same article that asserted documentary confirmation had hedged the identical claim three paragraphs earlier with the word "reportedly," contradicting itself on how well established it was.

What replaced it. The claim was removed from every article. In its place, articles now state what is documented: the Holy See purchased approximately 15,000 quintales of Puerto Rican coffee per year during the 1950s and 1960s, shipped by the Ponce export firm Sobrino de Mayol Hnos, making it the only government to purchase Puerto Rican coffee directly. A shipping document consigned to Vatican City is reproduced in Luis Pumarada O'Neill's La Industria Cafetalera de Puerto Rico: 1736-1969, published by Puerto Rico's State Historic Preservation Office.

Nineteenth-century papal and royal patronage is now described as tradition rather than record.

Articles affected: 16.

2. Conflicting Vatican supplier accounts — reconciled

What was published. Three different organisations were named as the Vatican's supplier across different articles: the Cooperativa de Cafeteros de Puerto Rico, Sobrino de Mayol Hnos, and the Café Rico brand — all citing the same 15,000-quintal figure.

What we found. The accounts were describing different stages of one trade. The Cooperativa, founded 1924, graded and prepared the coffee. Sobrino de Mayol Hnos of Ponce exported it. They were successive links in the same chain, not competing suppliers.

A dating point also emerged: Sobrino de Mayol was founded after Hurricane San Felipe in 1928, so the firm named in connection with the Vatican trade did not exist in the nineteenth century. The documented relationship is twentieth-century.

3. "Super Premium" described as an industry classification — corrected

What was published. Articles stated that only three coffees in the world carry a "super-premium" classification, naming a Puerto Rican brand alongside Jamaica Blue Mountain and Hawaiian Kona.

What we found. No coffee industry body issues such a designation. The Specialty Coffee Association grades coffee by cupping score, where 80 points and above qualifies as specialty; there is no recognised tier above it. The phrase is a brand name, and the claim traces to the brand owner's own product copy.

What replaced it. Articles now identify it as marketing language and explain the actual grading system.

4. A brand founding date presented as history — reframed

What was published. Articles stated as fact that a Puerto Rican coffee estate was founded in 1839.

What we found. Every published source for this date reproduces the same product description, word for word, from the brand owner's pages to retailer listings. No deed, property register, census return, survey, or scholarly reference could be located.

What replaced it. The date is now attributed to brand material rather than asserted, with the sourcing problem stated plainly. The date may be accurate; it is not documented.

5. Undisclosed common brand ownership — disclosed

What was published. Articles discussed several Puerto Rican "heritage" coffee brands as distinct estates from distinct regions, each with its own history. One article contrasted two of them against each other as opposites.

What we found. Alto Grande, Yauco Selecto, Yaucono, Café Rico and Café Rioja are all brands of Puerto Rico Coffee Roasters, a subsidiary of Coca-Cola Puerto Rico Bottlers.

What replaced it. Common ownership is now disclosed wherever it bears on a heritage claim.

6. Museum holdings asserted without examination — qualified

What was published. Articles stated that the Museo del Café in Ciales displays letters from royal households requesting Puerto Rican coffee and receipts from Vatican purchases.

What we found. This encyclopedia has not examined the collection. Published descriptions of its contents vary and could not be confirmed.

What replaced it. Articles now describe the museum as a repository of coffee trade material and state that its specific holdings have not been examined — a place to research rather than a claim already verified.

7. A quotation attributed to Theodore Roosevelt — removed

What was published. One article stated that President Theodore Roosevelt served Puerto Rican coffee at White House state dinners and called it "grand."

What we found. The only published source located for this is retail promotional material, including this encyclopedia's sponsor's own website. A quotation attributed to a named historical figure requires a real source.

What replaced it. The sentence was removed in full — both the quotation and the state-dinners claim.

8. Conflicting Hurricane San Ciriaco figures — reconciled

What was published. Articles variously gave the 1899 death toll as "approximately 3,000," the crop loss as "60%," and the storm's passage as "28 hours from east to west."

What we found.

The official death toll in Puerto Rico was 3,433; the National Hurricane Center's reanalysis gives approximately 3,369
Crop damage is properly stated as two figures: roughly 55 to 60 percent of plantings destroyed and about 90 percent of the harvest lost
The frequently cited "28" refers to 28 days of rain following the storm, not the duration of its passage. The eye crossed the island in about six hours, moving southeast to northwest

What replaced it. All figures now match across articles, with sources named.

9. A brand described as a legal appellation — corrected

What was published. One article described a commercial coffee brand as functioning like a wine appellation such as Bordeaux or Barolo, with "strict quality standards" that "guarantee" authenticity.

What we found. The brand is a registered mark exported since about 1990, drawn from particular estates rather than from all growers in the region. No standards document, certifying authority, or altitude threshold governing its use could be located.

What replaced it. The article now identifies it as a commercial brand and directs readers to Café de Puerto Rico, the island's actual denomination of origin, certified through the Oficina de Cafés de Puerto Rico under Ley 78-2019.

10. A brand dated a century and a half too early — corrected

What was published. An article stated that a named brand reached the Vatican and European royal tables during the 1830s to 1890s.

What we found. The brand has been exported only since about 1990.

What replaced it. The anachronism is now flagged explicitly rather than silently dropped, so readers encountering the claim elsewhere can recognise it.

11. Images not depicting their subjects — replaced

What was published. A review of every image on the affected articles, examining the files rather than their names, found extensive mismatches. Among them:

An article on an 1899 hurricane led with a satellite image of Puerto Rico's power grid during the 2017 Hurricane María blackout
An article on the nineteenth-century Yauco coffee estates led with a 1942 photograph of a Yauco shanty settlement, captioned as a hacienda landscape
An article on a Lares coffee brand led with a locator map of a San Juan suburb roughly 100 kilometres away
An article on Puerto Rico's Vatican trade illustrated its sections with a Venetian clock, a Polish porcelain cup, and an Antwerp salt cellar
A section on Puerto Rican neighbourhood shops was illustrated by a supermarket in the Netherlands
An image whose source record described Hurricane María damage in Puerto Rico was found, on inspection, to show a coffee estate in Karnataka, India

What replaced them. Period-appropriate material from the Library of Congress, the National Archives, the USDA, the Historic American Engineering Record, and Wikimedia Commons — each verified by examining the image rather than its filename.

12. A deceased pope described in the present tense — corrected

What was published. An article described Pope Francis's tastes in the present tense.

What we found. Pope Francis died on 21 April 2025 and was succeeded that May by Pope Leo XIV.

What replaced it. The passage is corrected. Articles also now note the coincidence that the sitting pope shares a name with Leo XIII, the pope named in the unsourced claim above, and that the two are separated by more than a century.

August 2026 — Citation Audit

A separate review examined every citation-shaped claim on the site: 128 claims across 36 articles, of which 38 were bibliographic — naming a journal, an author, a study year, or a participant count. All 121 articles were checked. The findings were serious enough to record in full.

13. Two studies cited that do not exist

What was published. The article on coffee and longevity cited a Mendelian randomization study attributed to the Karolinska Institutet and published in Nature Medicine, and a 2023 telomere study in the journal Antioxidants.

What we found. Neither could be located in the published literature. A search of PubMed returns no such Nature Medicine paper on coffee and mortality, and no coffee-telomere paper in Antioxidants for that year. Both citations appear to have been generated rather than drawn from a source.

There is a real Karolinska Mendelian randomization study — Yuan et al., Nutrients, 2021 — and it reached the opposite conclusion to the one the article reported. It used twelve genetic variants across fifteen cardiovascular outcomes and found that genetic predisposition to higher coffee consumption "was not associated with any of the 15 studied cardiovascular outcomes," concluding that previous observational studies "may have been confounded." The fabricated citation had inverted a null result into proof of causation, and had turned the study's fifteen outcomes into fifteen genetic variants.

What replaced it. The section was rewritten around the genuine literature. The article now reports that observational studies find a large and consistent association between coffee and longer life, that genetic studies designed to test causation have not confirmed it, and that some genetic evidence points in less favourable directions. The article's title was changed from "What 500,000-Person Studies Reveal" to "What the Large Cohort Studies Show," because no study cited reaches 500,000 participants.

This finding prompted a check of every citation across all 121 articles. The two fabrications were confined to this one page and did not spread.

14. A dose-response figure stated in the wrong unit

What was published. The article on coffee and heart health stated that each additional cup of coffee per day was associated with a 5 percent reduction in heart failure risk and a 6 percent reduction in stroke risk.

What we found. The source — a 2017 American Heart Association presentation — reports 7 percent and 8 percent per additional cup per week. Presenting a weekly figure as a daily one overstates the effect roughly sevenfold.

What replaced it. The corrected figures and the correct unit, together with a note that the finding was presented as a conference abstract rather than a peer-reviewed paper.

15. A meta-analysis misreported

What was published. That a 2019 meta-analysis drew on 21 prospective studies and found 21 percent lower all-cause mortality.

What we found. The study covered 40 studies and 3,852,651 subjects, and reported a relative risk of 0.85 — about 15 percent. The figure 21 appears to have been duplicated into both slots.

16. A pregnancy review attributed to the wrong author and its conclusion reversed

What was published. The article on coffee and pregnancy attributed an integrative review of 257 articles to "Francisco and colleagues" and described it as finding the evidence "generally supportive of moderate consumption being safe."

What we found. The review is by Rohweder et al., published in Reproductive Toxicology in December 2023. There is no author named Francisco. Its actual conclusion is materially less reassuring: the authors "found association between caffeine intake and pregnancy loss, low birth weight, cardiac and genital anomalies, higher body mass, and neurodevelopmental and neurobehavioral outcomes," and reported that in some studies, exposures below the 200-milligram guideline were associated with those outcomes.

What replaced it. The correct citation and the review's actual findings, quoted. The article now reports that the reviews genuinely disagree about whether a safe lower threshold has been established, rather than presenting the reassuring reading alone.

This was the most consequential single error found on the site. It appeared on a page read by pregnant women, and it erred in the reassuring direction.

17. Participant counts and institutional attributions corrected

Several studies were credited with the wrong participant numbers or the wrong institutions. A study of 498,134 people was described twice as "over 500,000." A 200,000-participant figure was credited to a paper that had 125,950. A study by researchers at the National Cancer Institute and Northwestern University was described as a Harvard finding. Claims of reduced cancer and diabetes mortality were attributed to a study that explicitly found no significant association with total cancer mortality. All have been corrected against the sources.

18. A brewing figure stated incorrectly, and attributed to a standards body

What was published. The article on coffee extraction stated that filter coffee should fall between 1.30 and 1.55 percent total dissolved solids, attributing this range to the Specialty Coffee Association. A range of 1.15 to 1.55 percent appeared in three further places on the same page.

What we found. SCA Standard 310-2021 specifies a brew strength of 11.5 to 13.5 grams per litre — 1.15 to 1.35 percent on the SCA Brewing Control Chart — from a solubles extraction yield of 18 to 22 percent. The figure 1.55 appears to be the top of the chart's strength axis mistaken for the top of its ideal region. No source supports it.

This mattered more than most corrections on this page. It was a practical figure a reader might use to adjust their own brewing, and it was attributed by name to a standards body that publishes a different number.

What replaced it. The correct range in all four locations, cited to the SCA standard and to a peer-reviewed paper that states the same figures. Readers should note that some older SCAA documentation gives 1.15 to 1.45 percent; that variation is recorded in the article's sources.

Also corrected on the same page. The research behind the Brewing Control Chart was attributed to the Massachusetts Institute of Technology with industry funding. Ernest Earl Lockhart was an MIT-trained biochemist, but he left MIT in 1955 to become the first research director of the Coffee Brewing Institute, and the brewing research was carried out there. The article also stated that the Specialty Coffee Association adopted the chart in the 1960s; the SCA was founded in 1982. Both have been corrected, and the article now states plainly that its primary source is a thirteen-page 1957 trade publication whose full text this encyclopedia has not read.

A note on how these errors arose

This encyclopedia's articles are drafted with AI assistance. Fabricated citations — plausible-looking references to studies that do not exist — are a documented failure mode of that technology, and the two found here fit the pattern exactly: real institution, real journal, real subject area, no such paper.

The response has been to verify every bibliographic citation on the site against its source, and to adopt a standing rule that no claim is published until the sentence supporting it has been located in the source named for it. That rule is recorded in the encyclopedia's Editorial Standards.

Readers who find a citation they cannot verify are asked to report it to Encyclopedia@PuertoRicoCoffeeShop.com.

Open Questions

Recorded here because they remain unresolved. Anyone able to close one is asked to write to Encyclopedia@PuertoRicoCoffeeShop.com.

The Museo del Café collection, Ciales. Its holdings have not been examined. Specific catalogued items relating to nineteenth-century royal or papal purchases would settle a question this encyclopedia currently leaves open.
Nineteenth-century royal and papal customers. No primary document naming an individual monarch or pope as a purchaser of Puerto Rican coffee has been located. Trading records, hacienda receipts, and family papers of the period survive in fragmentary condition.
The Mariani de-husking adaptation. The account that the Mariani family of Yauco adapted a cotton gin to de-husk coffee in the 1860s is consistently reported. No patent, period technical description, or scholarly treatment establishing the date or mechanism has been located.
Hacienda Santa Clara and Hacienda La Juanita. The Library of Congress record for the Yauco coffee estate at Barrio Río Prieto carries both designations. The relationship between them is unresolved here.
The 1839 founding date. Whether an estate operated at the stated location from that year is not established independently of brand marketing.
Current production figures. Published figures for Puerto Rican coffee production differ substantially between the 2022 Census of Agriculture and statements by Department of Agriculture officials. The two may measure different things. This encyclopedia reports both.

Last updated August 2026.