Dooley Scholarship 2016: Fill & Download for Free

Yes, they are very different. Actually, they are the opposite of each other, from any point of view. It doesn’t matter what you compare, you don’t find anything similar between them.Catherine's family is elegant, classy and well educated. In almost two decades, they rarely put a foot wrong.The only good thing you can say about Meghan's family is that her mother seems nice. The rest of them are a bunch of classless and bankrupt individuals. They can sell their daughter and sister for a few hundred dollars without batting an eyelash.CATHERINE, THE DUCHESS OF CAMBRIDGEThe family photo from the wedding of the Duke and Duchess of Cambridge (2011).The family photos from the christenings of Prince George (2013), Princess Charlotte (2015) and Prince Louis (2018).ParentsHer parents are married since 1980 and they have a very close relationship with Catherine and William (they attended William's graduation from Sandhurst; after George was born, William and Catherine spent the first few weeks at the new Middleton home and the grandparents are very involved with the children; the 3 families—Cambridges, Middletons and Matthews—spent multiple vacations together) and the Royal Family (they spent vacations with Prince Charles and the Queen at Balmoral, Christmas at Sandringham and were invited at the Ascot Races, family weddings and the Diamond Jubilee River Pageant).Thomas van Straubenzee with Michael, Catherine and Carole Middleton at Prince William’s graduation from Sandhurst in 2006.First official photo of Prince George taken by Michael Middleton, at Bucklebury Manor. George with his grandparents—with Michael on a yacht at the King’s Cup Regatta in 2019 and Carole at a playground in Berkshire, in 2015.Vacations—Ibiza, Spain in 2006; Trois Vallees, France in 2012; on their way to Mustique in 2012.Carole Middleton and the Queen at Balmoral in 2016.Carole, Michael and Pippa Middleton with James Matthews at Sandringham in 2017.Pippa and James Matthews at Sandringham in 2018.Michael and Carole Middleton at Royal Ascot in 2011.Michael and Carole Middleton at Royal Ascot in 2012.Michael and Carole Middleton with the Duke and Duchess of Cambridge at Royal Ascot in 2017.The Middleton and Matthews families at the weddings of Prince Harry in 2018 and Lady Gabriella Windsor in 2019.James Middleton with Pippa and James Matthews at Princess Eugenie’s wedding in 2018.The Middleton family at the Diamond Jubilee River Pageant in 2012.They are self-made millionaires and have a combined net worth of $70 Million. Their business, "Party Pieces" has an estimated value of $50 Million and they also own multiple properties and stakes in race horses shares.At the moment, they live at the $8 Million Bucklebury Manor, a Grade II listed Georgian mansion with an 18-acre estate.Bucklebury Manor in 2015.The estate has a tennis court where Prince George took classes from the family-friend Roger Federer. As big tennis fans, the Middletons are attending Wimbledon every year and are also playing tennis at the Queen's Club, in West Kensington.The Middleton Family attending Wimbledon in 2018 and 2017.Michael Francis MiddletonHe was born in 1949, in a wealthy family with ties to aristocracy and has three brothers: Richard (his son, Adam, is Princess Charlotte's godfather), Simon and Nicholas (his younger daughter, Lucy, is Prince Louis' godmother).Lucy Middleton (far left) and Adam Middleton (right) with the Duke and Duchess of Cambridge watching Cirque du Soleil at the Royal Albert Hall, on Catherine’s birthday, in 2013.Michael was educated at Clifton College, where he played rugby and tennis. He studied for six months at the British European Airways' Flight School to become a pilot, but switched to ground crew and worked as flight dispatcher for British Airways after graduating from the company's own internal course. Between 1984-1987, Michael was a manager for the British Airways in Amman, Jordan and by 1995 he gave up the job and focused on Party Pieces.Michael Middleton and his two daughters in Jordan, in 1986. Accompanied by Crown Prince Hussein, Prince William visited the same site, during his Royal Tour of Jordan, in 2018.His father was Captain Peter Middleton, an Oxford-educated RAF pilot who served in WWII. He was Prince Philip's co-pilot on a two-months tour of South America in 1962.Peter Middleton and Prince Philip, The Duke of Edinburgh in 1962.His mother was Valerie Glassborow, the daughter of a bank manager, who grew up in France. She and her twin sister, Mary, worked as code-breakers at Bletchley Park, during WWII. Valerie also served as a Voluntary Aid Detachment with the Red Cross.Catherine, the Duchess of Cambridge wears an Enigma brooch which belonged to her grandmother, Valerie, while visiting Bletchley Park in 2019.On the 150th anniversary of the British Red Cross, the Duchess of Cambridge shared this photo of Valerie (center) wearing her uniform during WWII.Michael's paternal grandmother was heiress Olive Middleton (née Lupton), the grand-daughter of Francis Lupton III (the Lupton Family was described as "landed gentry, a political and business dynasty") and great-grand-daughter of Elizabeth Martineau (the Martineau Family is “an intellectual, business and political dynasty”). She signed up as a volunteer with the Red Cross during WWI. Her brother, Lionel, studied at Trinity College (University of Cambridge) with Lady Diana's grandfather, Albert, 7th Earl Spencer and they joined together to fight in the WWI, where Lionel and his two other brothers were killed.London's National Portrait Gallery, holds nearly 20 portraits of ancestors of the Middleton Family.Carole Elizabeth Middleton (née Goldsmith)She is a former flight attendant turned businesswoman, born in 1955 to Dorothy Harrison, a clerical assistant and Ronald Goldsmith, a painter and decorator. Her family was coming from the working-class, her ancestors being coal miners and carpenters.Carole attended various state institutions and initially left school at 16, but soon returned and got four A-levels (art, economics, English literature and geography). She wanted to become a teacher, but her parents couldn't afford to sent her to university so she worked as secretary, while studying French and later got a job as ground staff at the British Airways. Carole established Party Pieces, in 1987, after returning to the United Kingdom.Her younger brother, Gary Goldsmith, was born in 1965. He is a multi-millionaire businessman who made his fortune from IT recruitment. During their dating years, William and Catherine spent some vacations at his villa in Ibiza and despite some controversial actions and the (mostly innocent) interviews he gave before 2011, he was invited at both of his nieces' wedding ceremonies. Gary went through three divorces (Miranda—1991/1996; Luan—1997/2006; Julia—2007/2009) and his current wife is Julie-Ann. He has a 17 years-old daughter, Tallulah Goldsmith, from his second marriage.Gary and Talullah Goldsmith attending the wedding of Pippa Middleton in 2017.SiblingsCatherine is very close to her siblings. After completing their studies, the three of them lived together in an apartment, in Chelsea. Pippa was the maid-of-honour and James read the lesson at their older sister's wedding.Philippa "Pippa" Charlotte Matthews (née Middleton)-Mrs. Matthews of Glen Affric the YoungerShe was born in 1983 and is a columnist, philantropist, businesswoman, author and socialite. Pippa attended the $51,000 a year Marlborough College and held a sports/all-rounder scholarship. She went on to graduate from the University of Edinburgh with a B.A. (Hons.) in English Literature. Here, she became friends with the son of the Duke of Roxburghe, Lord Edward "Ted" Innes-Kerr and the future Duke of Northumberland, George, Earl Percy.Following graduation, Pippa worked part-time for her parents' company, for a PR firm and as an event planer. In 2013, she set up PXM Enterprises Limited for which she is the sole director and shareholder. She is also a regular columnist for Waitrose Kitchen and wrote for The Spectator, Vanity Fair and The Sunday Telegraph.Pippa is ambassador for “Mary Hare School for Deaf Children” and “The British Heart Foundation” and she participated in numerous charitable sport events, like the Race Across America (a 5,000 km cycling race across the US), Bosphorus Cross-Continental Swimming Race (a 6.5 km swimming competition in Istanbul), London to Brighton Bike Ride and a 75 km swim-run competition in Sweden alongside her brother and her then-boyfriend, James Matthews, to raise money for the “Michael Matthews Foundation”, a charity established in honour of James' late brother, who passed away when he was descending Mt. Everest after becoming the youngest briton to reach the summit in 1999.In 2017, she married James Spencer Matthews in a $350,000 wedding ceremony which took place in Berkshire, with the reception being held at her family's estate.The Bucklebury estate before Pippa’s wedding.Pippa wore a bespoke gown ($10,000) by British fashion designer Giles Deacon with a veil ($2,000) from Stephen Jones and as accessories, she wore a pair of earrings ($13,000) created by Robinson Pelham, which were a gift from her parents with the ocassion of Catherine's wedding, the "Maidehair Fern" tiara from the same jeweller and a diamond hairpin to secure the veil. Pippa's wedding band ($20,000) is a Cartier creation and the engagement ring is rumored to be created by Robinson Pelham and it's believe to be worth more than $300,000.Pippa Middleton’s wedding in 2017.James is a British billionaire, former professional racing driver, hedge fund manager, and heir to the Scottish feudal title of Laird of Glen Affric, which makes Philippa the future Lady Glen Affric. His younger brother is TV personality Spencer Matthews (who is married to model Vogue Williams and James Middleton is the godfather of their son, Theodore) and his parents are David Mattews (who was previously married and has a daughter, Nina) and his second wife, Jane Parker, a Zimbabwe-born artist.Spencer Matthews and James Middleton at Spencer’s wedding in 2019.They are the owners of Eden Rock, an exclusive and prestigious resort in the Carribean, which was listed as one of the 100 best hotels in the world. Since 2007, they also own Glen Affric, a 10,000-acres estate in Scotland.James was educated at Uppingham School and spent most of his childhood at the family's residence called Caunton Manor (a 30-acre estate in Nottingham).The Matthews couple attending a wedding in Stockholm, in 2017.Pippa's and James' first son, Arthur Michael William Matthews, was born in the Autumn of 2018 and the family lives in a $22 Million five-story mansion in Chelsea, London.The home of the Matthews family in Chelsea, London.James William MiddletonHe is a businessman and was born in 1987. Like his older sisters, he also attended Marlborough College and later started a degree in Environmental Resources Management at the University of Edinburgh but left after a year.Inspired by his oldest sister' work in the mental health area, James wrote some articles and gave interviews on the theme, after revealing that he struggles with depression and attention deficit hyperactivity disorder.He owned some small businesses and at the moment he is working as "host" at Pippa's father in-law's Scottish estate.Like his other relatives, James is a sport lover, taking part in many events, like climbing Switzerland's Matterhorn peak, with his sister in Michael Matthews' honour. He is also a big dog lover.Pippa and James Middleton in Switzerland, in 2016.James Middleton and his dogs for The Telegraph in 2019.In October 2019, it was revealed that he is engaged to Alizee Thevenet, with whom he was first seen at the “Henry van Straubenzee Memorial Fund’s Christmas Carol Service”, the previous December.Alizee was born in France, speaks three languages (French, English and Spanish) and is a financial analyst, who studied accounting & finance in Brussels and has a master in investment & finance from Queen Mary University of London.Alizee Thevenet and James Middleton at Lady Gabriella Windsor’s wedding in 2019.They live together in an apartment in Battersea, London.MEGHAN, THE DUCHESS OF SUSSEXThe family photos from the wedding of the Duke and Duchess of Sussex (2018) and the christening of Master Archie Mountbatten-Windsor (2019).ParentsMeghan's parents are divorced. She has a close relationship with her mother, but is estranged from her father.Thomas Wayne Markle Sr.An awarded lighting director and director of photography, he was born in Pennsylvania, in 1945. His parents were Doris May Rita and Gordon Arnold Markle. Thomas has two brothers, Frederick and Michael.With his first wife, Roslyn Loveless, he has two children. They divorced in 1975 and he married for a second time to Doria Ragland, Meghan's mother, in 1979. This marriage also ended in divorce, in 1988. Thomas won $750,000 at California State Lottery in 1990, but he ended up filing for bankruptcy in 2016 over a debt of $30,000.He had a close relationship with Meghan, but since 2018, when it was revealed that he staged photos for paparazzi in return for money, he gave numerous interviews against his daughter to various TV shows and media outlets.Currently, he lives in Mexico.The home of Thomas Markle Sr. in Rosarito, Mexico.Doria RaglandShe was born in Ohio, in 1956 and her parents were nurse Jeanette Arnold and her second husband, Alvin Azell Ragland, who was working as an antique dealer. After her parents’ separation, her father, Alvin, remarried Ava Burrow, to whom Doria remained close after that marriage also ended in divorce. She has two older maternal half-siblings, Joseph Jr. and Saundra, and a younger paternal half-brother, Joffrey.Doria Ragland and Meghan Markle at the closing ceremony of the Toronto Invictus Games in 2017.After attending Fairfax High School, in Los Angeles, she worked as make-up artist. Following her divorce from Thomas Markle, she worked as a travel agent, yoga instructor and owned a small business, before filing for bankruptcy in the mid 2000s.Doria Ragland accompanies the Duchess of Sussex at the launching of her book in September, 2018.Doria completed a B.A. in Psychology and received a M.A. in Social Work from the University of Southern California. In 2015, she passed her social work licensing exam in California and worked as a social worker for a mental health clinic until 2018.She lives in View Park-Windsor Hills, Los Angeles, California.Doria Ragland’s home in California, USA.SiblingsMeghan is estranged from her half-siblings and both of them have publicly criticized the Duchess and like their father, they gave countless interviews to media outlets and various TV shows.Samantha Marie MarkleShe was born in 1964 and currently lives with her boyfriend, Mark Phillips, in Florida. Samantha is twice divorced and has 3 children: Noelle, Ashleigh and Christopher. She has been through personal bankruptcy, is estranged from her children, mother and brother and since 2008, when she was diagnosed with multiple sclerosis, she is using an wheelchair.Most famously, Samantha called Meghan a "shallow social climber". In October 2019, it was reported that she was under investigation following multiple allegations of cyberbullying.The (former?) home of Samantha Markle in Florida, USA.Thomas Wayne Markle Jr.He was born in 1966 and lives in Oregon with his fiancée, Darlene Blount.Thomas Markle Jr. and his fiancée, Darlene Blount.Like his older sister, he was twice married: first to Tracy Dooley with whom he has two sons, Thomas (who was arrested when he was wandering around Hollywood while high on drugs, shouting gibberish and only wearing a towel around his waist) and Tyler (who owns a cannabis business) and then, to Johannes Rawha.Thomas (first photo) and Tyler (second photo).He was arrested in January 2017 after threateningly holding a gun to Darlene's head during an argument while he was under the influence of alcohol. In July 2018 they were involved in another domestic violence incident.Thomas called Meghan "a jaded, shallow, conceited woman" and sent Prince Harry a letter, asking him to call off the wedding, because this is going to be "the biggest mistake in royal weddings history".The home of Thomas Marke Jr. in Oregon, USA.P.S. I am sorry for the grammatical mistakes.

Genealogical documentation.This answer is going to be very long. I have a lot of ground to cover, and a particular sense in which I treat documentation like an elitist snob. I also feel a need to begin with the idea of the whole of genealogy as an exercise in elitism. This may seem like an understandable (if not a logical) position for someone like me, who takes pride in his accomplishments as a researcher, but it’s not my position.I. One might start by thinking genealogy is a pretty snobbish subject to begin with. Why go even farther into the weeds to find an aspect of it to be snobbish about? Once people start doing it, though—no matter how privileged their upbringing, nor how distinguished their own accomplishments—they generally find that many of their ancestors weren’t very privileged at all. Just ask Matthew Stewart, for instance, who came from “old money” that actually originated with his great-grandfather (The 9.9 Percent Is the New American Aristocracy, The Atlantic, June 2018, part 9):I had long assumed that the Colonel was descended from a long line of colonels, each passing down his immense sense of entitlement to the next. … Robert W. Stewart was born in 1866 on a small farm in Iowa and raised on the early mornings and long hours of what Paul Henry Giddens, a historian of Standard Oil of Indiana [where Stewart was later an executive], politely describes as “very modest circumstances.” The neighbors, seeing that the rough-cut teenager had something special, pitched in to send him to tiny Coe College, in the meatpacking town of Cedar Rapids. It would be hard not to believe that the urgent need to win at everything was already driving the train when the scholarship boy arrived at Yale Law School a few years later.(However, the author’s views on modern-day inequality, expounded at length, are contested. See for example Jordan Weissman, Forget What the Atlantic Is Telling You. The 1 Percent Are Still the Problem, Slate, 18 May 2018.)Many of the sources that we find genealogically useful today are simply products of individuals’ efforts to make their way in life. They formed families, joined formal or informal organizations, traded in land and goods, and occasionally settled disputes in court or “out of doors.” In a modern society, all of these acts and more could produce records, whether the participants were “great,” humble, or somewhere in between. At its best, a genealogy is a collective biography of people of all classes, informed not only by records of the events that directly involved them, but also by a sensitive understanding of the social conditions under which the events unfolded. Elizabeth Shown Mills, “Genealogy in the Information Age: History’s New Frontier?” National Genealogical Society Quarterly 91 (2003): 260–77, at 277, conceives it as an answer to the fictional Martin J. Dooley of Chicago ([Finley Peter Dunne,] Observations of Mr. Dooley [1906/1902], 271, via Google Books).I know histhry isn’t thrue, Hinnessy, because it ain’t like what I see ivry day in Halsted Sthreet. If any wan comes along with a histhry iv Greece or Rome that’ll show me th’ people fightin’, gettin’ dhrunk, makin’ love, gettin’ married, owin’ th’ grocery man an’ bein’ without hard-coal, I’ll believe they was a Greece or Rome, but not befure.Not only is genealogy not a preoccupation with the aristocratic, as is sometimes alleged; a preoccupation with the aristocratic actually forecloses the full range of possibilities for genealogical study, and ultimately defeats its purposes, even when its subjects are the “greats” of the present day.II. No, my idea of elitist snobbery is not inherent in genealogy itself. It extends only to the nature and uses of the documentation that is proposed as evidence for family relationships. It’s not as if I haven’t been warned against it—or at least some version of it. This is precisely the purpose of Thomas W. Jones, Skillbuilding: Perils of Source Snobbery, OnBoard 18.2 (May 2012): 9, 10, 15, via Board for Certification of Genealogists.Jones begins with a distinction between sources that are preferred and those that are disdained. He proposes to base this distinction on “analytical tests”—evaluations of documents that focus on whether or not their contents are likely to be factually correct in every detail. However, Jones does not dwell on the content of these tests. He seems to suppose that genealogists apply the criteria, not to individual records, but to whole categories of documents. The most basic criterion of evaluation is whether or not the records appear, on their face, to reflect eyewitness testimony of events as they happened. Thus Jones’s examples of typical preferred sources include “most original court, immigration, land, military, probate, religious, tax, and vital records.” On the other hand, disdained sources “typically are derivative sources, often poorly documented, containing information from hearsay or an unknown person, like many online family trees.”It is somewhat surprising that Jones, writing for the membership of one of genealogy’s foremost professional groups, does not address the character of the information as primary, secondary or undetermined, nor the character of the evidence as direct, indirect, or negative. Even his depiction of sources as original or derivative is incidental. Yet all three dimensions—source, information, and evidence—are fundamental to Mills’s Evidence Analysis Process Map (Elizabeth Shown Mills, Evidence Explained: Citing History Sources from Artifacts to Cyberspace, 3rd ed. [Baltimore: Genealogical Publishing Company, 2015], front endpaper, emphases hers):Basic Principle: SOURCES provide INFORMATION from which we identify EVIDENCE for ANALYSIS.However, the mystery is dispelled when we remember that neither information nor evidence is judged according to the character of the source. “As a rule,” Mills explains, “primary information carries more weight than secondary, although either class of informants can err. Moreover, any statement can represent both firsthand and secondhand knowledge. … Therefore, each piece of information within a source must be appraised separately” (Evidence Explained, 25). Jones appears to be suggesting that source snobbery consists of assigning or withholding preference to sources, irrespective of their contents (“information”) and interpretation (“evidence”). If someone does that with a source, then not only are the information and evidence not important to the evaluation; it is even hard to maintain that the evaluation is based on an “analytical test” or any other objective qualification.Under the evidence analysis process that genealogists had inherited from academic history, it was possible to speak of a source as a unity, and as primary or secondary in its entirety. This conception lasted long enough to leave its mark on Donn Devine, “Evidence Analysis,” Professional Genealogy: A Manual for Researchers, Writers, Editors, Lecturers, and Librarians [1st ed.], ed. Elizabeth Shown Mills (Baltimore: Genealogical Publishing Company, 2001), 327–42, esp. 333. At this point, Devine seemed to conceive of an original source as a document created near the time of the events recorded, by bearers of primary information. It was not universally recognized at the time that original records can and often do bear secondary information; that there is no time limit on the bearing of eyewitness testimony, short of the witness’s demise; and that different pieces of information within the same source might not bear the same weight. (I would guess that since 2001 Devine, who like Jones and Mills is a longtime associate of the Board for Certification of Genealogists, has adjusted to the Board’s position on the nature of an original record. This is succinctly expressed in Genealogy Standards, 50th Anniversary ed. [Nashville, Tenn.: Ancestry.com, 2014], 77–8 and passim.)One way to understand Jones’s distinction between preferred and disdained sources is as a revival of the old dichotomy between primary and secondary sources, under different names. On this conception, genealogists have a tendency to prefer some sources because they are treatable as primary, and to disdain others that are treatable as secondary. The instinct behind this practice is much the same one that gave rise to the old dichotomy in the first place. Unfortunately, the practice relies on obsolete standards of evidence analysis. Jones’s point is that sources of preferred kinds (“primary sources”) may contain errors, and sources of disdained kinds (“secondary sources”) may actually be accurate. Thus it is hazardous to assume that a source is accurate or inaccurate merely from a presumption for or against every source of the same type.III. Everything in Jones’s essay is true as far as it goes. It is often helpful for an eager researcher to remember that the involvement of people concerned in a record is not a guarantee of its accuracy. But what Jones doesn’t tell us about when we can trust a typically disdained source is practically everything.Jones proposes that a disdained source is exactly like a preferred one, in terms of whether it is validated by the sum of the evidence bearing on a given research question. Any information from any source must be correlated with information from other sources before any judgment can be made on its accuracy. Jones concludes, “effective family historians consult and assess all sources, regardless of type, that might help answer their research questions. They exclude no potentially useful source, and they trust no unverified source.”Jones’s Table 1 shows ten instances from nine separate articles in the National Genealogical Society Quarterly from 1995 through 2010, in which a preferred source was found to contain an error. By comparison, Jones strains to name instances in which disdained sources were used, in his own words, to “show the only or most efficient path to reliable, informative records, or … provide evidence critical to the researcher’s conclusion.” His footnote flag at the end of this sentence cites two additional articles from the same journal. In addition, among the references describing preferred sources with erroneous information, there is a third essay whose title indicates that disdained sources were also considered there, under different terminology.George L. Findlen, “Using Questionable Sources Productively: The Parents of Rial, Edwin, and George Plummer of Alna, Maine,” National Genealogical Society Quarterly 96 (2008): 165–76. However, as we shall soon see, its interest arises from much more than the author’s use of “anonymous undocumented queries written more than a century after Plummer brothers were born,” which is all that Jones’s footnote mentions.Thomas W. Jones, “The Three Identities of Charles D. McLain of Muskegon, Michigan,” National Genealogical Society Quarterly 96 (2008): 101–20 at 107.Thomas W. Jones, “The Children of Calvin Snell: Primary versus Secondary Evidence,” National Genealogical Society Quarterly 83 (1995): 17–31.These cases have particular bearing on what it means to decide a genealogical case based on a source that might be considered questionable on its face. They require closer examination.IV. We begin with Findlen. That the three Plummer men were brothers, and two of them were natives of Alna, was documented by conventional methods. Users of an electronic database accessed by Findlen had reported parents for Edwin and Rial, but this claim could not be documented in multiple collections of local records from Alna and adjacent towns.The breakthrough came when Findlen inspected two letters from one Alna-area historian and genealogist—Henry Otis Thayer—to another—William Davis Patterson. Thayer wrote on behalf of “A correspondent” (that is, someone who had previously written to him), giving something resembling the same names for George’s parents that later were given for Edwin and Rial’s, and requesting confirmation of the names of George’s father and grandfather.Because none of the other correspondence survives—not from Patterson, nor from Thayer’s source—it could not be determined exactly what Thayer had learned from the others. Nevertheless, Patterson was closely related to the Plummer family—a grandnephew of the Plummer brothers’ mother—and, in his personal files, recorded all three brothers as sons of the same proposed parents.Perhaps Patterson incorporated information relayed to him by Thayer and his informant, but he also could have relied on any number of local eyewitnesses as well, who were contemporaries of his Plummer cousins as well as himself. Curiously, Findlen does not mention these last possibilities. He rates all of the information in all of the sources as secondary, on the grounds that neither writer had specified any eyewitness or documentary basis for naming the parents of Rial, Edwin, and George. Findlen’s rationale for the identification lies in the statements’ mutual corroboration and apparently independent origins.In truth, there actually is very little reason to believe that the findings of Thayer and Patterson were wholly independent of one another. There also is no reason to believe that each didn’t draw on additional sources—Patterson in particular being, Findlen concedes, “as well-positioned and qualified as anyone in the early twentieth century to identify the Plummer brothers’ parents” (p. 173). His record is surely the one that is most likely to have drawn on the widest range of relevant surviving testimony at the time, and the one that best sustains a “preference.” Thayer may give some of the same information, but even Jones, in “Perils of Source Snobbery,” admits that merely duplicating information from one manuscript to another does not constitute corroboration. (“For example, a family Bible’s birth date chiseled onto a gravestone remains uncorroborated.”) Thayer’s facts might well have been copied twice—from Patterson’s notes to his letter and then from Patterson’s letter to Thayer’s. For all the attention Findlen gave to his sources’ claims to authority, he seems to have quailed at the thought that any of the claims might have been superior to the others.One has to ask, finally, what Findlen would do without Patterson. He would be reliant, by default, on Thayer. His second letter suggests that Patterson might have already reported the correct answer, but Thayer was not quite ready to accept it. By Jones’s standard of evidence analysis, this would leave Findlen in quite a fix. He would have to consider whether the case could be decided on one manuscript after all. Or, he could simply confess that this is very close to the state of the case with Patterson and Thayer together—because, on balance, Thayer gives little reason not to assign Patterson the priority.Neither do the twentieth-century queries—which, contrary to Jones’s characterization, were not anonymous at all. They were signed by their contributor, George E. Plummer of Berkeley, California, whom Findlen identifies as George’s great-grandson. The range of topics that George E. engaged over the time span in which the queries were printed shows that he acquired relevant knowledge as the series progressed. However, no other correspondence to or from him is available, and we once again have no means of determining his informants’ names—or what facts he might have obtained, directly or indirectly, from Thayer and Patterson.As with Thayer, it is impossible to rule out contributions by Patterson in George E. Plummer’s writing; and because of this, we once again have no choice but to evaluate Patterson’s information on its own terms. That Patterson acquits himself remarkably well is not an index of what Thayer, Plummer, and Findlen might have learned independently of Patterson, because Findlen ultimately has not demonstrated that there is any such outside source on which to draw.V. Jones cites his McLain article, like Findlen, for a very limited purpose: as “an example of undocumented family trees on Ancestry.com as key to solving a complex genealogical problem.” In fact, this abbreviated reference serves his purpose in “Perils of Source Snobbery” better than his reference to Findlen.Jones had not succeeded in establishing the parents of Charles D. McLain by conventional research methods. He had found a possible, post-divorce second marriage for him, but no direct proof that it pertained to the same man—although he found the circumstantial evidence in favor compelling enough to pursue. So he searched for entries of this wife among the online family trees. According to an entry contributed by DeWayne G. Baker, one David R. McLain had entered the same marriage. Baker also identified his parents. Conventional methods of research on this family, now focused on David and his parents as well as Charles, produced further circumstantial evidence in favor of identifying David with the same man who had entered both of the marriages.Jones makes it quite easy to see how the Ancestry search marked a turning point in his research. However, had the narrative centered on Charles’s life rather than on the process employed by Jones, it could just as easily have been written without using Ancestry. Throughout the rest of the article, Jones does not refer to it again.The essential point is that Jones somehow obtained a statement that David had been credited with the same 1886 marriage that Jones himself was prepared to attribute to Charles. The medium on which he obtained it is novel, but the technique is not. An undocumented statement literally handwritten on paper might have roused him to the same mission, which he would have tried to fulfill with the same procedure.Without Jones’s subsequent research, of course, this purpose could never have been realized. But now that we have means of sustaining the identification that are better than referring to Baker, with his lack of documentation, we need never refer to Baker again—as an authority on this claim or any other. In other words, having used something “as a hint” entails no future obligation to regard it as a reliable source on other genealogical statements. This is part of the bargain: what it means to say, once again with Mills, that “each piece of information within a source must be appraised separately” (Evidence Explained, 25).If one undocumented statement in a source of overall doubtful accuracy has been useful for research, it still leaves a genealogist free to disregard all the other statements in the same source. How’s that for snobbery?VI. The title of the Snell article suggests it was written under the lingering influence of the old dichotomy between “primary sources” and “secondary sources.” Before Elizabeth Shown Mills had even printed Evidence! Citation & Analysis for the Family Historian (Baltimore: Genealogical Publishing Company, 1997), let alone devised the conceptual distinction among sources, information, and evidence at the center of her Evidence Analysis Process, this borrowing from the tradition of academic history was understandable. Still, a question beckons. Did Jones’s view of “secondary evidence” in 1995 correspond with his view of “disdained sources” in 2012?His summary of the preferred source in “Perils of Source Snobbery” describes it thus: “Will omitting testator’s eleven children and falsely identifying three heirs.” As to the latter, Calvin Snell’s will suggests that there were two sons and one daughter who are not thus named elsewhere. So much for “primary evidence.”What “secondary evidence” does he raise up in opposition to it? It consists of four separate post facto writings that each name incomplete sets of the children of Calvin Snell. One is a compiled four-volume Snell manuscript, dated 1934; one is a 1979 memorandum by Calvin’s great-granddaughter, copied from notes dictated to her by her mother, Calvin’s granddaughter; one is a genealogy of a son-in-law’s family, printed in 1896; and one is a loosely organized manuscript of uncertain date and authorship. He introduces them in the text as “unofficial sources” (p. 20), but summarizes them in a two-page table whose heading calls them “Secondary Sources” (p. 22). Now, of course, they are more properly described as “sources bearing secondary or undetermined information,” exactly the type that drove all the other studies I have described. This one meshes seamlessly with the others.Jones’s sources do not agree on every name, but there is considerable consistency across the lists. As with Findlen, Jones takes this consistency as evidence that they were composed separately from independent precursors. Jones has more justification for his use of the procedure; it is much easier to rule out the participation of one informant in the creation of multiple records in the Snell case than the Plummer case, and the Snell sources all contradict Calvin’s will in similar ways.Even so, it is still possible to rank them by preference, choosing any criteria that we like. Suppose that I rank them according to the likelihood that each source had immediate contact with Calvin or his children. Personally, on these grounds I would first take the 1896 source, then the 1934, then the 1979, then the undated. If any grandchildren of Calvin can be proven to have lived through 1934 or had contact or correspondence with the author of the Snell genealogy, that alone could put its reliability at a par with the 1896.Anyone may disagree with my ranking. The point is that when we get down to making empirical judgments—especially in cases where our information is contradictory—we need defensible reasons for selecting one apprehension of the facts over another. We may even refrain from making a judgment if none of the available apprehensions of the facts seem to be satisfactory. In the process, we sometimes rate one source more highly than another, even if they disagree minimally over the facts. In a conflict, we are not debarred from declining to accept a source because of factual claims that resist validation, and we are not entitled to believe things solely for self-ratifying reasons.VII. Whether you have had the forbearance to read all of the above or not, I feel that a summary is in order. Genealogy is already mistaken by many for a preoccupation with the elite. But researchers don’t accomplish anything worthwhile if they bring that attitude to the topic, and in any case I doubt that many people with a sophisticated grasp of the subject consider it a channel for elitism and snobbery.What I have, instead, is a qualified dissent on Thomas W. Jones in “Perils of Source Snobbery,” particularly if it entails a presumption that one source is never to be preferred over another. I have no brief against his claim that a piece of information cannot be excluded on the mere grounds that it arises from a derivative or authored source of doubtful accuracy, nor that standard original records do not always provide adequate proof of basic facts. In principle, there is nothing objectionable about these claims.However, the proposition that pursuing identifying statements in non-standard sources can overcome the limitations or complete lack of relevant original records requires a more powerful defense than Jones offers in the essay, simply because it seems so unbelievable on its face. Jones does cite a few cases to offer support for this claim, without defending it at length. Unfortunately, Jones overstates George Findlen’s reliance on a twentieth-century researcher’s queries, and in addition Findlen himself seems to overstate his source writers’ absence of coordinated effort.Jones’s own McLain and Snell cases are more informative on the procedure of developing an acceptable statement of probable facts from a combination of derivative records and corroborating statements from original records. However, even his cases do not foreclose the possibility of ranking different derivative sources according to singular criteria, and judging the strength of a statement’s probability according to how much confidence it seems to warrant compared with other sources touching on the same question.I certainly don’t expect a research project beginning with a collection of documents, each apparently unacceptable on its face, to end by affirming their probable accuracy on the sole grounds that they happen to agree on more points than they disagree. I am reminded of the literary critic Frederick Crews in another context, speaking of “the classic sophistry of … ascribing a cumulative weight to reports that, when regarded one by one, are lighter than air” (“The Mind Snatchers” [1998], Follies of the Wise: Dissenting Essays [Emeryville, Calif.: Shoemaker & Hoard, 2006], 204). In every successful instance under review, there are always independent documentary grounds for believing that the genealogical claims are accurate. Deciding the truth of genealogical claims on these grounds whenever possible, and withholding judgment on “disdained” sources as long such evidence remains unavailable, I take as essential precepts of responsible source snobbery.If, as appears likely in the Plummer case, the main sources of information all appear to have developed from one spring, there is no need—and considerable risk—to posit the “independent” later creations of records whose freedom from influence of the main spring cannot be verified. We may have no option but to accept that there is only one truly original record for a given statement of fact. In that case, the search for corroborating evidence aims only to show that this record pertains to, and gives reliable information on, the person whose biography and family relationships are the objects of the study.Some of the ideas above, especially as they relate to Professional Genealogy (2001), I had also shared in July 2016 with other members of ProGen 29. They bear no responsibility for my adaptation here.

Absolute pitch (AP) is the ability to identify and name the pitch of a sound without external reference. Often, accuracy and speed at naming isolated musical pitches are correlated with demographic, biological, and acoustical parameters to gain insight into the genesis and evolution of this ability in specific cohorts. However, the majority of those studies were conducted in North America, Europe, or Asia. To fill this gap, here we investigated the pitch-naming performance in a large population of Brazilian conservatory musicians (N = 200). As previously shown, we found that the population performance was rather a continuum than an “all-or-none” ability. By comparing the observed distribution of correct responses to a theoretical binomial distribution, we estimated the prevalence of AP as being 18% amongst regular music students. High accuracy thresholds (e.g., 85% of correct responses) yielded a prevalence of 4%, suggesting that AP might have been underestimated in previous reports. Irrespective of the threshold used, AP prevalence was higher in musicians who started their musical practice and formal musical education early in life. Finally, we compared the performance of those music students (average proficiency group) with another group of students selected to take part in the conservatory orchestra (high proficiency group, N = 30). Interestingly, the prevalence of AP was higher in the latter in comparison to the former group. In addition, even when the response was incorrect, the mean absolute deviation from the correct response was smaller in the high proficiency group compared to the average proficiency group (Glass's Δ: 0.5). Taken together, our results show that the prevalence of AP in Brazilian students is similar to other non-tonal language populations, although this measure is highly dependent on the scoring threshold used. Despite corroborating that early involvement with musical practice and formal education can foster AP ability, the present data suggest that music proficiency may also play an important role in AP expression.IntroductionEvolution of auditory and vocal systems has reached high levels of complexity in humans, allowing the emergence of both speech and music (Patel, 2003). In this respect, one important feature of sound is pitch, which grants prosody to spoken speech and melody to music. Although, pitch perception is ubiquitous in the animal kingdom (Weisman et al., 2004), some individuals are able to identify and name the pitch of a particular sound without any external reference, a supposedly rare ability known as absolute pitch (AP) or perfect pitch (Bachem, 1937; Takeuchi and Hulse, 1993; Levitin and Rogers, 2005; Deutsch, 2013).The prevalence of AP is highly heterogeneous and culture-specific, as it relies on how signifiers (pitch label) and signified (periodicity of sound waves) relates to each other (Nattiez, 1990). Considering Western populations in Europe and North America, the prevalence of AP has been estimated between 0.01% (Bachem, 1955) and 0.07% (Profita and Bidder, 1988). However, the methodology of these early studies is hard to replicate and include biased samples and subjective measures. It is now acknowledged that the prevalence of AP can be as high as 75% among specific cohorts, as in music schools and conservatories (Baharloo et al., 1998; Gregersen, 1999; Deutsch et al., 2006, 2009; Miyazaki et al., 2012; see also Sergeant and Vraka, 2014).It has been proposed that contrasting AP prevalence among different populations could be explained by the influence of genes (Baharloo et al., 1998; Gregersen, 1999; Theusch and Gitschier, 2011) and/or environmental constrains (Baharloo et al., 1998; Miyazaki et al., 2012; Wilson et al., 2012; Deutsch, 2013). In this respect, different hypotheses were proposed to explain AP development and expression (for a review, see Takeuchi and Hulse, 1993; Deutsch, 2013). The first hypothesis suggests a correlation between genes and the cognitive ability to listen to pitches in absolute terms, initially put forward by pedigree analysis that found a higher prevalence of AP among first-degree relatives (Baharloo et al., 1998; Profita and Bidder, 1988), siblings (Baharloo et al., 2000), and twins (Gregersen, 1998). The higher prevalence among Asian in comparison to Caucasian subjects was also considered as an indirect evidence for the genetic predisposition for AP (Gregersen et al., 2001), although cultural factors cannot be ruled out (see the second hypothesis below). Despite these evidences, sophisticated whole-genome linkage analyzes failed to reveal specific AP-related genes, which prompts the conclusion that AP may be genetically heterogeneous (Theusch et al., 2009). Therefore, pinpointing specific genes related to the predisposition to AP is now seen with skepticism, and alternative hypotheses need to be considered.A second hypothesis proposes that full acquisition of AP ability can only take place when subjects are taught to relate pitch and musical labels within the “sensitive period” of development (Deutsch, 2013). The early learning hypothesisproposes that internalization of sound and its association with a specific label relies on the increased plasticity governing the formation of neural assemblies early in life (Deutsch, 2013). In fact, AP prevalence is higher among musicians who started their formal musical education before 6 years old (Baharloo et al., 1998; Deutsch et al., 2006; Miyazaki et al., 2012). Nonetheless, some studies have reported improved performance after extensive training in both children (Crozier, 1997; Russo et al., 2003; Miyazaki and Ogawa, 2006) and adult populations (Vianello and Evans, 1968; Brady, 1970; Heller and Auerbach, 1972; Van Hedger et al., 2015).Finally, the third hypothesis considers the musical training regime and/or environment. For example, it has been shown that the prevalence of AP can be higher among musicians who speak tonal languages in comparison to non-tonal language speakers (Deutsch et al., 2006, 2009, 2013). In this case, the cognitive demand to extract meaning from similar sounds with varied intonations would extrapolate to other tasks, and therefore, contribute to pitch recognition and naming. However, early music education policies in Asian countries may also contribute to the differences in the observed prevalence (Miyazaki and Ogawa, 2006; Miyazaki et al., 2012), as music education methods (i.e., “fixed-do” vs. “moveable-do”) and the type of instrument used in learning classes (Miyazaki et al., 2012; Deutsch et al., 2013). In this respect, at least to our knowledge, few studies have attempted to investigate the influence of music proficiency in the expression of AP, which may have been overlooked in previous reports.Music is one of the most important aspects of Brazilian identity. Some rhythmic and harmonic styles of the Brazilian music, like choro, samba, and bossa-nova, are well-known for its originality and diversity (McGowan and Pessanha, 1998). Interestingly, most Brazilian musicians start their music education informally (Feichas, 2010) and governmental support is scarce. Together, those conditions restrict the access to formal educational institutions to the most “gifted” musicians. In the present study, we used a pitch-naming task (Deutsch et al., 2006) in a large population of conservatory students to determine the prevalence of AP in a cohort of Brazilian musicians. Early studies using self-reported questionnaires observed an AP prevalence of 5% among Brazilian music students (Germano et al., 2011, 2013). However, these results could be biased since musicians may covet this ability, and to overcome such limitation, psychophysical tests are mandatory (Bermudez and Zatorre, 2009). The methodologies of psychophysical tests often yield varied degrees of performance at the population level raising another relevant question: is there a clear threshold to separate AP possessors from non-AP possessors? Curiously, this threshold is set arbitrarily in most studies. Here, we try to overcome the threshold problem by statistically comparing individual performance in a pitch-naming test with a theoretical binomial distribution. This approach allowed the identification of AP possessors whose performance surpassed chance. For comparison purposes, we also included a sub maximal threshold used in previous reports (Deutsch et al., 2006, 2009, 2013). Finally, we analyze and present our results in light of aforementioned hypothesis—characterizing our sample according to (1) the age of onset of musical training and formal musical education, (2) musical instrument, and (3) musical proficiency. This third factor was evaluated by comparing the prevalence of AP among regular music students (average proficiency group) and those selected to take part in the conservatory orchestra (high proficiency group). To our knowledge, this is the first empirical study aimed to determine the prevalence and characteristics of AP in Latin America.Materials and MethodsParticipantsStudents from the Music School of the Federal University of Rio Grande do Norte were approached in the beginning of the academic year (February and March 2016) and were invited to take part in an experiment related to music perception. Thirteen out of 41 courses allowed the application of the test. A total of 200 regular students (55 female, 24 ± 7 years-old; 145 male, 25 ± 9 years-old) satisfied the inclusion criteria (reported no auditory or neurological disorders) and fully completed all stages of the test. Participants did not receive any financial support (as dictated by Brazilian regulations) and provided written informed consent before the experiments were realized, in compliance with international standards. All protocols and procedures were previously approved by the Ethics Committee of the Federal University of Rio Grande do Norte (CEP-UFRN #31273114.5.0000.5537).To further investigate the role of musical proficiency in the prevalence of AP, a second group of high proficiency music students were included in the test (N = 30; 4 female, 21 ± 4 years-old; 26 male, 24 ± 5 years-old). This group represents especially talented students who have succeeded at a high-level orchestral audition, in which the musicians perform musical pieces belonging to the orchestral repertoire. The audition is judged by teachers from the conservatory, as well as by an expert in the musician's instrument. Besides the social visibility within the conservatory and the positive impact in their careers, the selected students are entitled to receive scholarships. For these reasons, the competition to fulfill these positions is high and only high proficiency musicians are selected.QuestionnaireBefore performing the pitch-naming task, participants were asked to complete a questionnaire informing their age, history of musical education (start and duration of both informal and formal musical training), training habits (schedule and estimated amount of time practicing), and instruments played. These questions aimed to investigate the contribution of musical education during sensitive periods.StimuliThe set of stimuli consisted of 36 piano tones (1 s duration), ranging from C3 (131 Hz) to B5 (988 Hz), synthetically generated (GarageBand, Apple Inc.). The harmonic sequence consisted of the fundamental frequency and its harmonics conferring a musical piano timbre to the synthetic sound. The stimuli were presented through two loudspeakers (~70 dB) with the assistance of a personal computer running SuperLab 4.5 software (Cedrus Corporation, Inc.).Pitch Identification TestThe pitch-naming task used was modified from a previous report (Deutsch et al., 2006). A set of 36 stimuli was presented randomly and repeated once. Each set consisted of three blocks of 12 stimuli, with an interval of 30 s between blocks to minimize fatigue. Before starting the test, we presented a training block with four trials. Participants were asked to write down the name of each note as fast as possible. The interval between stimuli was set to 4.25 s. To avoid the use of strategies based on relative pitch judgments, the tonal distance between each tone sequentially presented was higher than one octave, and never had one semitone interval (not considering the octave). No octave information was requested and no feedback was given, neither during the training, nor during the test blocks. The tests were conducted in the classrooms and in the conservatory theater for the regular and orchestra groups, respectively.Data Analysis and StatisticsThe percentage of correct responses was calculated by scoring the exact identification of pitch. Thus, no semitone errors were allowed as even small deviations in pitch identification violate the general meaning of AP. Besides the criterion of pitch distance, AP prevalence also depends on the threshold used to infer the percentage of AP possessors. We formed two different criteria for AP, due to the lack of consensus in the literature (Gregersen, 1999; Athos et al., 2007; Miyazaki et al., 2012; Deutsch et al., 2013). The first one, defined as conservative AP (cAP), included subjects scoring at least 85% of correct answers. This threshold was based on previous publications (see Deutsch et al., 2009, 2006) and it was selected to allow comparisons between the present work and similar previous large-scale studies. The second threshold considers the identification of AP possessors in the population from a statistical point of view (Sergeant and Vraka, 2014). Thus, the chance expectation can be determined by the binomial distribution, which calculates the likelihood Pkof observing k correct choices in n number of trials, given the probability p in each trial, according to the equation:Pk=(n!k!(n−k)!)pk (1−p)n−k[math]Pk=(n!k!(n-k)!)pk (1-p)n-k[/math]Considering 72 trials and the probability of correct response in each trial to be 1:12, the mean probability of correct responses was calculated as 6 correct responses (i.e., 8.3%). We used the Clopper–Pearson method (Clopper and Pearson, 1934; binofit function in Matlab) with 99% confidence interval to determine the threshold of significance to regard the performance above chance. Here, the first integer value outside this confidence interval was 15 (i.e., 20%), and this value was considered the threshold for our inclusive AP (iAP) approach. Therefore, participants with 15 or more correct responses, but less than 61 (i.e., 85%) were considered as presenting some level of absolute pitch ability, and treated separately from the cAP group.Mean absolute deviation (MAD), with and without correct trials, was computed and correlated with the performance (correct responses yields absolute deviation = 0, while one semitone mistake yields absolute deviation = 1). Correlation between the first and second block was calculated to control for changes in performance during the test (Bermudez and Zatorre, 2009). Correct responses and MAD were used to compare the effects of pitch class and pitch label. Differences in frequency distributions of categorical variables were compared using Pearson's chi-squared test. We used one-way analysis of variance (ANOVA), followed by multiple comparison post-hoc Bonferroni test, to compare independent variables (age, amount of daily hours practicing with musical instruments in the last year, mean absolute deviation, and performance). Welch's F-test was used in case the homogeneity of variance assumption was violated. In some comparisons, effect size was calculated using Glass's Δ method. Data are presented as mean ± standard deviation, and a p-value smaller than 0.05 was considered of statistical significance in all comparisons made.ResultsPrevalence of AP AbilityThe pitch-naming test revealed that performance varied significantly in our cohort of music students (Figure 1). It also reiterated that the population performance is a continuum rather than a discrete bimodal ability (Figure 1A). This poses the challenging decision in determining the threshold to classify a participant as AP possessor, which in turn, influences the prevalence of the ability. Using the threshold of 85% of exactly correct responses (Deutsch et al., 2006), we observed a prevalence of 4% (cAP group). The inclusive criteria (iAP), based on a binomial distribution, instead yielded a prevalence of 14% (Figure 1B). Together, cAP and iAP groups made 18% of our cohort (35 out of 200 subjects) while the others (non-AP; nAP) performed similarly to chance. The prevalence of AP using the self-reported questionnaire returned similar values (3.5%) to the cAP group, but surprisingly the participants who reported having the ability were not the same of those who surpassed the threshold (conservative and inclusive) on the pitch identification test (Table 1). The false-positive and false-negative rates were 2.5% (4/162) and 90% (29/32), respectively, when considering the self-reported questionnaire as the predictive condition. It is important to note that most of the false-negative subjects belong to the iAP group (24/29). No significant differences were observed in the prevalence of AP among male and female participants (Male: 19% and Female: 15%; p > 0.05, chi-square test), age at the time of the test [in years-old; nAP: 26 ± 9, iAP: 22 ± 6 and cAP: 24 ± 7; F(2, 197)= 2.79, p > 0.05, one-way ANOVA], and the amount of daily hours practicing with musical instruments in the last year [in hours; nAP: 2.5 ± 1.7, iAP: 2.9 ± 1.5, and cAP: 2.8 ± 2.3; F(2, 197)= 0.64, p > 0.05, one-way ANOVA].FIGURE 1Figure 1. Quantification of Absolute Pitch (AP) demonstrates that population performance is a continuum rather than an “all-or-none” ability. (A) Ranked performance (each bar represents one subject) shows that subjects with low performance identify white-key notes with more accuracy than black-key notes. Right of the graph shows the representation of the conservative (cAP) and inclusive (iAP) thresholds used for separating populations. Insets represent the histogram of the distances from correct responses to the presented stimuli. (B) The chance of correctly naming the pitch was calculated using a binomial distribution (solid line). The histogram represents the observed performance distribution. The dashed line represents the 99% confidence interval for chance-level scores (35 subjects, 18%, were above threshold). (C) Mean absolute deviation (MAD, in semitones) for each of the three groups. (D) Scatter plot showing the correlation between percentage of correct responses and mean absolute deviation. Note that the observed data do not fit in shaded area (simulated data), mainly for the iAP group. (E) Mean absolute deviation of incorrect responses. Note that iAP group responded closer to correct pitch than the nAP group. (F) Consistent responses of iAP and cAP were verified by correlating the mean absolute deviation between the first and the second blocks of 36 trials. *p < 0.05, t-test.TABLE 1Table 1. Contingency table of self-reported AP ability per group of psychophysically identified AP possessors or non-AP possessors.To determine whether the iAP group constituted a different population from the nAP group, we compared the mean absolute deviation between groups. The smallest and highest mean absolute deviation values were observed for the cAP and nAP groups, respectively, while iAP group showed intermediate values [F(2, 18.53)= 1047.30, p < 0.001; Welch's F-test; Glass's estimator of effect size between cAP and iAP = 2.03, and between iAP and nAP = 5.28; Figure 1C]. However, this was not surprising since mean absolute deviation and percentage of correct responses—the variable that we used to classify the groups—are strongly correlated (R2= 0.90, p < 0.001; Figure 1D). To better understand the distribution of the mean absolute deviation in our cohort, we simulated the values of the mean absolute deviations (using uniform distributions) for each possible performance value (200 iterations per score). The gray-shaded area in Figure 1D shows the simulated data and how our sample downwardly dispersed from it. This observation suggests that as the performance increases, the pitch distance between the response and stimulus decays faster. To test whether the distances of responses to the stimuli were indeed smaller in the iAP group in comparison to the nAP group, we recalculated the mean absolute deviations excluding the correct responses (i.e., absolute deviation = 0; Figure 1E). This approach allowed the demonstration that iAP group indeed differs from nAP group, once the former guessed more closely the stimulus pitch even when making wrong choices [F(2, 13.24)= 37.58; p < 0.05; Welch's F-test; Glass's estimator of effect size between cAP and iAP = 0.55, and between iAP and nAP = 3.46]. Finally, we tested the consistency of the responses for the AP groups by correlating the mean absolute deviations in the first and in the second blocks of stimuli (Figure 1F). We observed a positive correlation between the blocks (R2= 0.69, p < 0.001; Pearson test), although most of the subjects (mainly from the cAP group) improved their performance during the test, as revealed by an asymmetrical distribution of points above and below the diagonal line (which represents the equal performance between blocks; Figure 1F).Contribution of Pitch Class to the PerformancePerformance for natural (a.k.a., white-key) and accidental (a.k.a., black-key) notes varied in our population, mainly in the iAP group (Figure 2A). Interestingly, nAP group, whose performance was below chance, clearly fetched more white-key notes than black-key notes [Figure 2B; t(164)= 16.45, p < 0.001, paired t-test; Figure 2C; chi-square for observed vs. expected responses: χ2(1, 11752)[math]χ(1, 11752)2[/math] = 3216, p < 0.001]. Similar unbalanced performance for white-key notes was also observed in iAP group [t(26)= 7.30, p < 0.001, paired t-test; Figure 2C; chi-square for observed vs. expected responses: χ2(1, 1942)[math]χ(1, 1942)2[/math] = 160, p < 0.001]. However, no difference between white-key and black-key notes was observed in the cAP group.FIGURE 2Figure 2. Participants of nAP and iAP groups identified white-key notes better than black-key notes. (A) Performance of each pitch chroma for each group of participants. The performances for black-key notes (sharp/flat) are represented in a gray background. (B) Averaged correct responses according to pitch class (i.e., white-key and black-key notes). In the nAP and iAP groups, the performance for white-key notes was higher than for black-key notes *p < 0.001, paired-t-test. (C) Averaged white-key and black-key notes responses, irrespective of whether they were correct or incorrect. Dashed lines represent the expected white-key and black-key notes responses according to the set stimuli. *p < 0.001, chi-square test.Onset of Musical Training and Formal Musical EducationTo investigate the relationship between AP and the early onset of musical training, as previously reported in the literature, we divided participants in bins of age (years-old) of onset in both informal (<7: N = 19, 8–11: N = 46, 12–15: N = 98, 16–19: N = 28, 20–23: N = 4) and formal (<7: N = 4, 8–11: N = 27, 12–15: N = 58, 16–19: N = 60, 20–23: N = 22) musical practice. We observed a higher prevalence of AP in musicians who started their musical practice early in life (Figure 3). This was true when considering both the onset of musical practice (Figure 3A) and formal musical education (Figure 3B). While the latter probably involves classes of musical theory and some sort of organized methodology (not investigated here), the former deals with informal and probably exclusively “playing by ear” approach of learning pitch.FIGURE 3Figure 3. Absolute pitch has a higher prevalence in musicians who started musical training early in life. Percentages of AP possessors according to the (A) age of onset (in years) of musical practice, and (B) age of onset (in years) of formal music education. Solid lines represent subjects of the cAP group, and dashed lines represent subjects of the iAP group.Instrument ClassSince it has been suggested that early piano lessons could increase the prevalence of AP (Miyazaki, 1989), we compared the performance (Figure 4A) and the mean absolute deviation (Figure 4B) according to student's instrument modality. One-way ANOVA showed no differences in either parameters [performance: F(5, 194)= 1.2; p = 0.33; mean absolute deviation: F(5, 194)= 1.6; p = 0.16].FIGURE 4Figure 4. Pitch identification performance does not depend on the instrument played by the subjects. (A) Performance and (B) mean absolute deviation. Box plots of the distribution scores represent the median (horizontal line), the interquartile range (box width), and the outliers (gray crosses). The black dots represent the mean. Str, strings; Key, keyboards; Wind, winds; Perc, percussions; Voic, voice, and Mult, multi-instruments.Increased AP Prevalence among High Proficiency MusiciansTo examine how music proficiency influences the prevalence of AP at the population level, we performed the same experiment in a subpopulation of the music conservatory. This second experiment included musicians who were selected to be part of the conservatory orchestra (see Section Materials and Methods). The same thresholds used to separate the regular students were applied here. We observed a higher prevalence of AP in the high proficiency group (cAP = 7%, iAP = 40%) in comparison to the average proficiency group (cAP = 4%, iAP = 14%, Figure 5A). Yet, both groups of music students did not differ in their musical history or recent daily practice, as shown by the onset of musical practice [t(47)= 0.29, p = 0.77, t-test], the onset of formal instruction [t(47)= 0.09, p = 0.92, t-test], and the number of hours in their daily practice in the last year [t(47)= 0.82, p = 0.41, t-test; Figure 5B]. Excluding the correct responses (i.e., absolute deviation = 0), students from high proficiency group guessed more closely the stimuli than average proficiency group, as revealed by the smaller mean absolute deviation of the high proficiency group [t(45)= 2.02, p < 0.05, t-test; Figure 5C]. The effect size was estimated to be 0.5 (Glass's Δ method).FIGURE 5Figure 5. Orchestra musicians presented a higher prevalence of AP and an improved performance in pitch-naming task. (A) The prevalence of AP was higher in orchestra participants (higher proficiency) in comparison to regular students (average proficiency) of a music conservatory, irrespective of the threshold used to identify AP (i.e., iAP or cAP). (B) The onset of musical practice and the onset of formal education did not differ between both populations, as the informed averaged hours of study per day in the last year. (C) Mean absolute deviation of correct responses between average proficiency and high proficiency groups. *p < 0.05.DiscussionHere we present the first quantitative investigation of AP ability in Latin America using traditional psychophysical methods. Previous work, which only employed self-report questionnaires to assess AP, described a prevalence of 5% among Brazilian conservatory students (Germano et al., 2011, 2013). This prevalence is similar to the one described here using self-reported questionnaire (3.5%; Table 1). However, some individuals who reported being AP possessors did not reach the threshold criteria in the pitch-naming test (4 out of 7). Anecdotal reports suggest that musicians covet AP ability, although the reasons for such aspiration are undefined. AP has been associated with both improved musical transcription capability (Dooley and Deutsch, 2010) and impaired interval recognition (Miyazaki, 1995). A possible crave for AP could have an important impact on the conclusions driven from studies using self-reported questionnaires exclusively (Gregersen et al., 2001). Here, participants knew they would take part in a pitch-naming test immediately after answering the questionnaire, which might have attenuated the false-positive rate and boosted the false-negative rate. In this report, we showed that estimations based on self-reported questionnaires and pitch identification test evaluation can differ significantly, which makes the use of standardizing psychophysical tests mandatory, as previously suggested (Bermudez and Zatorre, 2009).Few papers have reported the pitch-naming performance distribution in the population (Athos et al., 2007; Bermudez and Zatorre, 2009; Miyazaki et al., 2012). In those that do, it is apparent that the performance in the population is actually a continuum rather than a discrete bi-modal attribute, although there may be a tendency for the formation of clusters at the extremities depending on the applied method. For example, using a web-based test Athos et al. (2007)found a distribution with a bimodal tendency; however, the lack of supervision during the test may be crucial, because there is no guarantee that participants did not adopt any reference before each stimuli block. This would allow the use of other skills, such as relative pitch, or even the use of an instrument that can give some feedback to the participant, which would increase participants' performance, especially those with intermediate performances. Another factor that can influence participants' performance is the way they are recruited, as early discussed by Bermudez and Zatorre (2009). In this sense, we used a sampling method that does not allow the self-selection and that, despite being applied in groups, occurred under the supervision of experimenters, ensuring that there is no external reference available to the participants—an indispensable condition for measuring AP.One key aspect when investigating the prevalence of AP is the inclusion criteria and the threshold used to determine whether the participant has the ability. In most studies, only a percentage of correct responses in a given number of trials combined with an arbitrary threshold to identify AP possessors is used (Deutsch et al., 2006). Conservative studies define a threshold spanning from 85 to 100% of correct responses (Miyazaki, 1989, 1990; Deutsch et al., 2009, 2006), while inclusive approaches use smaller thresholds. For example, some authors would consider evidence of AP when performance is simply above chance (Oechslin et al., 2010). As expected, this approach yields a higher number of individuals with AP. Also, some studies allow one or even two semitone distance from the presented stimuli as a correct response (Keenan et al., 2001; Loui et al., 2011). Such approaches include various levels of AP, which decreases the possibility of identifying and distinguishing different types of AP. In the present work, we applied a validated pitch identification test (Deutsch et al., 2006) in the classroom of the music conservatory, and observed a prevalence of 4% (conservative threshold) or 18% (inclusive and conservative thresholds summed; Figures 1A,B). These values are in accordance with previous findings in conservatory and occidental populations (Baharloo et al., 1998; Deutsch et al., 2006; Miyazaki et al., 2012). Together, these results demonstrate that AP is not as rare as previously suggested (Bachem, 1955; Takeuchi and Hulse, 1993), when considering a population of music students.To further investigate whether the iAP group is in fact a set of some type of AP possessors, we analyzed the mean absolute deviation of incorrect answers (Figure 1E) and it revealed that the iAP group is more accurate in their responses than the nAP group (which responded as expected by chance). This result, together with the statistical method used here for the detection of AP possessors, reinforce the idea that, despite not presenting high performance in pitch identification task, the iAP group has some ability of absolute pitch perception.Pitch Class and the Label Preference EffectMost studies that addressed the performance according to pitch class were done in AP possessors only, although subject's classification and their levels of performance varies significantly between studies. Regardless, all reports reviewed here found that white-key notes were more accurately and faster identified than black-key notes (Miyazaki, 1988, 1989; Takeuchi and Hulse, 1991; Athos et al., 2007; Bermudez and Zatorre, 2009). In our study, no significant difference in white-key vs. black-key notes performance was found in cAP group; however, a clear white-key note preference was observed in both iAP and nAP groups (Figure 2). Although, the reasons for pitch class preference in iAP group can be speculated, it would be difficult to extend this explanation for the nAP group, whose performance was below chance. We believe that pitch class preference can be explained by several factors. One of them is the perceptual magnet effect as suggested earlier by Athos et al. (2007). The hypothesis behind this concept states that the inability to distinguish similar sounds as belonging to different categories would result in naming auditory stimuli into one single prototypic label. In fact, nAP and iAP groups were more likely to write down a white-key than a black-key note when compared to the expected white-key and black-key proportion in the chromatic scale (Figure 2C). Another factor is the leaning and readiness for the participant to write down notes without mistake, which increases the likelihood of making correct choices by chance. Additionally, the specific note with the highest performance in nAP group was the “C,” which turns to be the reference pitch for most students when they start learning the major scale. All these factors argued previously, combined with the fact that the pitch-class effect was present in nAP group (Figure 2A), reveal that non-acoustic features, as label familiarity (e.g., historical reference notes) and pitch symbolic representation (e.g., presence or absence of accidents—sharp or flat), can influence pitch performance. This suggests that the road toward true (genuine) AP ability involves refraining cultural pitch preferences.Factors That Influence Absolute Pitch AbilityThere is evidence suggesting that playing a musical instrument, especially in the initial years of musical training, the sensitive age (Vitouch, 2003; Deutsch et al., 2009, 2006, 2013), may influence the development of AP (Miyazaki, 1989, 1990; Miyazaki and Ogawa, 2006). We tested this hypothesis in our work, but found no significant result when we measured the performance of the volunteers in the pitch-naming test in relation to the modality of the instrument (Figure 4). It has been suggested that the sensitive period would span from 3 to 7 years old (Takeuchi and Hulse, 1993; Baharloo et al., 1998; Deutsch et al., 2009), and in the present work, we observed that the earlier a musician started practicing music, the higher was the prevalence of AP (Figure 3). However, the age of onset referred here is twice the one described in previous publications (Miyazaki et al., 2012; Deutsch et al., 2013). One possible explanation for such discrepancy may be related to cultural differences, where Brazilian music students often start to learn and practice music by playing “by ear” with friends and/or family instead of formal enrollment at a music school. Interestingly, the prevalence of AP rose significantly when considering the onset of formal music education (Figure 3B). Finally, intriguing evidence suggests that early training may be neither necessary nor sufficient to the full expression of AP ability (Brown et al., 2003).Considering the higher age of onset of musical training in our sample of AP possessors (in comparison to previous reports), one may ask whether the neurobiology of the AP reported here is qualitatively different from the AP developed during the sensitive period. Early and recent studies have proposed the existence of different strategies of AP perception, named pseudo-absolute pitch and quasi-absolute pitch (Bachem, 1937; for a review see Levitin and Rogers, 2005). In those cases, linking pitch memory to pitch labeling would be less automated and more cognitively demanding than in individuals who acquired the ability early in life. Future studies are needed to determine whether different strategies yielding absolute pitch perception in fact exist. Possible methods should attempt to categorize the performance of participants using varied forms of stimuli (e.g., musical and non-musical sounds with different spectral statistics), while taking into account other physiological markers, such as reaction time, heart rate, skin conductance, movement-related EMG and EEG evoked potential (Itoh et al., 2005).Language has also been proposed to influence musical abilities (Elmer et al., 2013, 2014) and the development and expression of AP (Rakowski and Miyazaki, 2007; Deutsch, 2013). It has been suggested that linguistic experience early in life can alter phonetic perception (Kuhl et al., 1992), and such environmental constrain would facilitate the development of AP ability (Deutsch et al., 2004). It is generally assumed that Portuguese comprehension does not depend on tonal interpretation (Marques et al., 2007), despite regional accents, which can vary significantly in different parts of the world to produce rich melodic speech (Fernandes, 2007). In addition, prosody can play an important role in social communication adding complex layers of interpretation to the daily conversation; but prosody does not define the meaning of any spoken word when in isolation, as it does for tonal languages (Yip, 2002). Therefore, we believe that the melodious accent of the Brazilian language did not contribute to the prevalence of AP among Brazilian musicians. In this respect, AP prevalence was found to be higher in China, a tonal language country, in comparison to United States conservatories (Deutsch et al., 2006). This result was later confirmed by another study that observed a higher prevalence of AP only among students who speak a tonal language fluently in one music school in the United States (Deutsch et al., 2009). These observations are in agreement with a previous non-controlled, questionnaire-based survey, which found high prevalence of AP among Asian students (Gregersen, 1999). However, another large-scale experiment with similar methodology compared the prevalence in music schools in Japan and Poland, and found a higher prevalence in Japanese students-who do not speak a tonal language (Miyazaki et al., 2012). The authors argue that learning methodologies can better explain the differences in AP prevalence, which raises the question about the role of music proficiency in the expression of AP.The Influence of Music ProficiencyMusicality can be hard to define (Sloboda, 2008), despite bona fide attempts to create a framework capable of measuring it (Seashore, 1915; Levitin, 2012). Musical assessment tests appraise patches of music proficiency in the real world and consequently, scores can be highly dependent on the performance in specific features, as perception (time and melody) or sensory-motor integration (sight-reading). Thus, these tests do not provide the opportunity for musicians to exhibit their proficiency as a whole, and therefore, important aspects of musicality are neglect, for example, creativity, expression, or emotional communication (Levitin, 2012). Here, to investigate the relationship between proficiency and absolute pitch ability, we compared two populations of musicians of the same school, the regular students and those selected to take part in the Music School Symphony Orchestra. As mentioned early, those students comprise a special group of musicians who are selected by a tough orchestral competition, in which various aspects related to music proficiency are evaluated, such as execution, interpretation, technique, and sight-reading.In the orchestra group, we observed that AP prevalence was higher (cAP = 6.7%, iAP = 40%) than in the average proficiency group (cAP = 4%, iAP = 13.5%; Figure 5A). This result suggests that musical proficiency is associated with increased performance in a pitch-naming test. However, this result is neither explained by the onset of musical training or formal music education, nor by how much participants practice music daily. Moreover, it is not possible to assert causality between proficiency and AP prevalence. In the same way that intense training could lead to higher pitch identification performance, AP possessors could become high performance musicians easily. Further studies are necessary to unveil whether musical proficiency leads to AP, or AP leads to higher proficiency, or both phenomena are consequences of a common cause. Several studies showed correlations between musical proficiency and other complex abilities, such as speech perception (Schön and François, 2011; Christiner and Reiterer, 2013) and mathematical performance (Schmithorst and Holland, 2004). Interestingly, in all these studies the correlations are associated with an enhanced working memory. Not surprisingly, there is evidence that AP is associated with an increased capacity of auditory digit span, i.e., the ability to memorize and recall a specific sequence of numbers (Deutsch and Dooley, 2013). In this way, it was recently shown that individuals with high working memory capacity are better in learning/improving discrimination of pitch categories (Van Hedger et al., 2015). It is possible that an increased working memory capacity is the common link between musical and other non-musical abilities, in the same way that the prevalence of AP and higher musical proficiency seem connected. Still, further studies are needed to disentangle the participation of music proficiency and increased capacity of working memory in AP development.ConclusionsUsing a combination of questionnaire and a pitch-naming task, we were able to quantitatively measure AP prevalence in a population of Brazilian musicians for the first time. In addition, we demonstrated a white-key preference over black-key notes in subjects without AP ability, which suggests that this preference is not related to acoustic properties of pitch. Despite our volunteers started musical practice later than reported in previous studies, it was possible to confirm that there is a relationship between the prevalence of AP and the age of onset of musical practice. This suggests that the interaction between plasticity and training is more complicated than the simple opening and closing window of the sensitive period. Thus, we speculate that probably there are many different ways to achieve some degree of absolute pitch ability. Furthermore, using a more natural method of measuring musical proficiency (admission to a fine symphony orchestra instead of the standard quantitative protocols) and a pitch identification test, we identified an association between musical proficiency and AP. However, further studies are necessary to unveil the role of different variables that may contribute to this correlation, such as working memory.Author ContributionsRL and CQ conceived and designed the experiments and analyzed the data. RL and SM performed the experiments. RL, SM, and CQ wrote the manuscript.Conflict of Interest StatementThe authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.AcknowledgmentsThe authors would like to thank Katarina Leão, João Bacelo, Ernesto Soares, and Diego Laplagne for their valuable comments in the early versions of this manuscript, and the staff of the Music School of the Federal University of Rio Grande do Norte for technical support during data collection. Research supported by funding from CNPq, CAPES, FINEP, and FAPERN. RL was a recipient of CAPES scholarship.