I have AJP Crown to thank for this link: “Surnames of Occupation,” by C.M. Matthews (from History Today 13: 1963). It begins “Modern English surnames are so many fragments of medieval conversation, crystallized into permanent form,” and is full of interesting tidbits. A couple of excerpts:

Naturally, we must begin with Smith, the commonest English surname of any type. Among occupational names, it is in a class by itself with 5,750 examples, more than twice as many as any other. The reason for this multiplicity is not so much that metal-workers were numerous as that they were important and widespread. On the skill of the smith, both rich and poor depended for the most essential things of life, the tools of husbandry and the weapons of hunting and war. Every community in the land must have one, every castle, every manor; and so distinctive was his trade that he would seldom need another name.
Everyone knows that Smith comes first; but probably not many could name the second among names of occupation. Two names run very close together, Clark and Taylor, the former being slightly ahead with 2,740 as against 2,570. It is not at all surprising that the clerk should be so high on the list. Like the smith, he had a skill that was needed everywhere and rare enough to be valued—literacy. […]
It is natural that there should have been a miller in every village, and that, like the smith, he should take his name from his trade. Its numbers are harder to assess than some, because it has survived in various forms: Milner, the Early Middle English form which generally developed into Miller, and—with exactly the same meaning—Millward or Millard. These names taken together amount to 1,710.
But, besides these, there are 780 more closely connected with the building from which the trade was inseparable. Anglo-Saxon “mylen” gives us Milne, Mill and, even more commonly, Mills (the final “s” attached to many common names remains a puzzle to the experts). Many of these names must have referred to the miller himself, but some to his servants or merely to people who lived at or near the mill. If we add all these, the milling business would nearly equal the tailoring, but still remain in fourth place on our list.

I hadn’t realized that final -s in names was a mystery, but now that I think of it it’s certainly not clear where it would come from.


  1. Curious that there are so few Shoemakers, while (as far as I can tell) Schumacher is a common enough name in Germany. I wonder how the names tally compares to other european countries with occupation-derived surnames.

  2. Sutter?

  3. I remember once reading the suggestion that the final “s” implied “son”. The idea was that “son” was appended in Scotland and oop North – thus Richardson, Patterson, Nicholson and so on. Whereas in the Midlands, only “s” was used – thus Richards, Peters, Nicholls ….
    The Welsh took up the surname habit from their Midlands English neighbours, my source argued, thus Williams, Evans, Johns, ……

  4. J ap Crown would therefore have the alternative J Crowns.

  5. @s/o: What I’ve read is that shoemaking as a distinct full-time occupation is very late medieval, and surnames were in general adopted later in Germany than in England. Most if not all Americans named Shoemaker had an ancestor named Schumacher.

  6. There is also Shuster and Shubert (shoe-wright). I don’t know what the relative frequencies are in Germany.

  7. David Marjanović says

    I bet Mills is the mill’s man. There are places in Germany where the surnames of immigrants were formed by forming the genitive of the place they came from – I know a John J. Wiens.
    Information on the frequency and geographic distribution of German surnames is plentiful and should be easy to find.

  8. As one Dr. Oliphant said (and was quoted by Mencken in The American Language): “Many a Pennsylvania Carpenter bearing a name that is English, from the French, from the Latin, and there a Celtic loan-word in origin, is neither English, nor French, nor Latin, nor Celt, but an original German Zimmermann.

  9. Information on the frequency and geographic distribution of German surnames is plentiful and should be easy to find.
    Here. Just enter a surname.

  10. des von bladet says

    At work we have a “Schoenmakers” (presumably of Midlands ancestry) and until recently also a “Schoenmaker”. I bet some of their and their namesakes’ ancestors also have American descendents…

  11. Des: Might not they be of Dutch origin?

  12. John, once times rubs enough detail off a name it can be nearly impossible to determine it it’s Englisih or Dutch or deutsch and the family itself may not know.
    There are lot of Smiths and Millers in the US whose ancestrs were never near England. Unless you know what to look for “Snyder” is not going to appear very German at first glance.
    This is more than the normal Anglicization that turns 乐/樂 into “Locke” or 里,力,离,利,李,立,礼,栗,励 or 厉 into “Lee.” This is more like dialect leveling.

  13. Why isn’t the s simply a genetive? Mills is “of the mill”, as in serving it or living next to it.

  14. Has anybody noticed the large number of French surnames that end in -ec or -ac? Ditto for place names. Seems quite disproportionate to the number of nouns with those endings.
    Related: A fairly common first name in France is Patrick. What’s that ‘k’ doing there? French allows Patrice, as in Saint-Patrice, although it seems to prefer Patrick d’Irlande. French had no trouble developing Dominique from Dominicus, and it did produce Patrique from Patricius, but it seems nobody likes it much.
    According to this site, a mere 89 male babies in France were given the name Patrique between 1946 and 2009, while 137,599 male babies were given the name Patrice in those years, and 394,665 male babies were given the name Patrick in those years.

  15. des von bladet says

    John: I work in the Netherlands, so yes.

  16. I once wrote a letter — a polite rejection of an article submitted to an academic journal — to someone named Dominique. Years later I found out that he was a man (I had thought of “Dominique” as a woman’s name), and only then did I realize how offensively condescending the letter had been. It taught me a lesson about the sexism that lurks in the best-intentioned of us.

  17. (I had thought of “Dominique” as a woman’s name)
    A common error among Anglo-Saxons, I think, and for many years I (like, I presume, many others) entirely misunderstood the song for that reason.

  18. Yes, me too! But then I also got the singing nun mixed up with the flying nun.

  19. Athel Cornish-Bowden says

    Has anybody noticed the large number of French surnames that end in -ec or -ac? Ditto for place names. Seems quite disproportionate to the number of nouns with those endings.
    I don’t know about -ec (maybe Breton?), but I associate -ac with Acquitaine, and I think that is where the -ac comes from.
    On the other hand, as Mr Hat has remarked, Smith is by far the most common English surname, whereas Fabre is not particularly common in France, and Schmidt seems to be somewhat less common in Germany than Smith is in England. Yet the numbers of smiths of all kinds must have been similar in the three countries. Conversely, there are lots of Duponts in France, but not so many Bridges in England.
    Incidentally is a good source of information about this kind of thing. For example, “Hatt” occurred at about 30 in a million surnames in the UK in 1998. (No Hats, however, apparently.)

  20. Athel Cornish-Bowden says

    (I had thought of “Dominique” as a woman’s name)
    It often is, of course. I know Dominiques of both sexes. DSK’s manliness, however, seems not to be in question.

  21. Has anybody noticed the large number of French surnames that end in -ec or -ac?
    Where’s marie-lucie when we need her?

  22. My neighbors is Moscow were the Смит’s, literally Smiths, but it wasn’t anglicized Dutch or anything. As I recall, it was a great-grandfather nom-de-revolution, and he was nicknamed like this because his stint as a new immigrant in New York City.
    I read that surname selections in Russia have a strong regional variation, with Smith-equivalent being the most popular near Tula, and Parsons-equivalent, near Arkhangel?

  23. “Parsons” is just the English equivalent of McPherson. Anyway, what other names ending in “son” or “sons” don’t mean “son of”?

  24. Fabre is not particularly common
    Well, there’s Lefebvre / Lefèvre.
    Elsewhere, Kovács, Kowalski, Кузнецов, Ferrari.

  25. what other names ending in “son” or “sons” don’t mean “son of”?
    Tyson. Gerson.

  26. What I’ve read is that shoemaking as a distinct full-time occupation is very late medieval, and surnames were in general adopted later in Germany than in England. Most if not all Americans named Shoemaker had an ancestor named Schumacher.

  27. The U.S. Census provides lists of surnames and given names sorted by decreasing frequency in the U.S. population:
    surnames from the 2000 census (zip archive with CSV and XLS formats)
    surnames from the 1990 census
    Hispanic surnames from the 1990 census (PDF)
    male given names from the 2000 census
    female given names from the 2000 census
    2000 documentation (PDF)
    1990 documentation

  28. “Parsons” is just the English equivalent of McPherson
    But McPherson still means the same occupation, the priests?

  29. @Dmitry: yes, so I’ve always been told.

  30. -ac is in fact the Gascon reflex of the common Gaulish-derived placename emding -acum. Parisian -y, Swiss -ey, etc. I’ve seen a map.

  31. Paul Ogden, Athel Cornish-Bowden, Hat: I’m not Marie-Lucie, but I may be able to shed some light on French place-names and surnames in -AC.
    Basically, -AC is the reflex, in the langue d’oc (including Gascon) and Lower Brittany, of a latinized Gaulish suffix, -ACUM, which was attached to surnames and indicating a person’s property. In the langue d’oïl area this suffix has -Y as its reflex.
    Thus, in the days of Roman Gaul a landowner by the name of SABINIUS would own a property called SABINIACUM, and if this property was located in what is today the langue d’oïl area its name will be “Sévigny” today, whereas in Lower Brittany or the langue d’oc area its name will be “Sévignac” (There will be plenty of variants in both instances). From these place-names derives its use as a surname, of course.
    The fact that you find -AC in Lower Brittany as well as Southern France doesn’t mean there is any kind of special relationship between those two areas within France. It’s partly a shared archaism, partly a coincidence: when Celtic speakers settled Lower Brittany from Cornwall in the fifth century they borrowed the place-names from the local Romance speakers, who at the time, like Romance speakers everywhere in Gaul, had -ACO or -AGO as their reflex of -ACUM: it later became -AC in both Lower Brittany and Southern France, with Northern France innovating far more radically, phonologically, and turning the suffix into -Y.
    And indeed there is an -AC/-Y line, involving place-names in Brittany, which separates the two linguistic zones of (pre-modern) Brittany. In the (Western) area (where -AC dominates) Breton was the vernacular language. In the Eastern area (where-Y dominates) the langue d’oïl (more specifically: Gallo) was the vernacular language.
    Hope this clarifies things a little.

  32. Parson is a variant of person, and the parson is so named because he was in law Latin persona ecclesiae, the human mask on top of an impersonal institution, almost the first institution known to English law (the distinction between Queen and Crown, for example, is very modern and is even now informal: it is the Queen, not the Crown, who prosecutes most criminals, despite the name of the Crown Prosecution Service).

  33. The use of -person to replace -man in compounds and avoid alleged sexist connotations is first recorded 1971 (in chairperson)
    Since I heard so many times that -person is genderist too, & ought to be replaced by – peroffspring, I must have accepted too readily the notion that there is SON in perSON 🙂 Glad to see how misguided that was!

  34. Trond Engen says

    Where’s marie-lucie when we need her?
    I know the answer to that!

  35. Etienne: I may be able to shed some light on French place-names and surnames in -AC.
    You certainly have! Thank you.
    Athel Cornish-Bowden: there are lots of Duponts in France, but not so many Bridges in England.
    Bridgestone, the huge Japanese tire and auto parts manufacturer, at one time marketed motorcycles in North America under the name Rockford. According to Wiki, the name Bridgestone is a “calque translation and transposition of ishibashi, meaning ‘stone bridge’ in Japanese.” The founder’s name was Shojiro Ishibashi.

  36. Wouldn’t Briggs be a name from bridge? “Brig” certainly means bridge where I grew up.

  37. i’m with @Yuval with -s as “of/pertaining to”.
    in yiddish bynames/surnames (which predate, but persisted after, state-enforced christian-style heritable surnames were imposed over jewish-style patronymics), it’s pretty common to find names like “dovid soyres” /david sarah’s husband/ which use that construction.

  38. Trond Engen says

    Dearie me: Wouldn’t Briggs be a name from bridge? “Brig” certainly means bridge where I grew up.
    Did you also say e.g. mig for midge and heg for hedge?

  39. Wouldn’t Briggs be a name from bridge?
    This site thinks so and adds that Brigham is a related name.

  40. Trond Engen says

    I see that DSL has neither mig nor heg. But it does list a riddle compound heg-beg “nettle” of equally obscure etymology. I can see “hedge” as the first element, but what is beg? Is there a big “barley”? Oh, yes. So let me suggest a kenning “the barley of the field’s edge”.
    Oh, ege “edge” is there. Also, rig “ridge”. And now I want to derive big from ON byggja v. “build”. I’m just not sure how.

  41. ON bygg is ‘barley’; byggja is ‘inhabit’. Not sure they’re related.

  42. Trond Engen says

    ON byggva (not byggja, sorry) didn’t just mean “build” but even “imagine the physical appearance of”. Furthermore, ON bygg “barley” is probably essentially the root of the verb, pointing to a meaning “make grow, appear”. Well, I’m imagining an adjective *bygg “prone to grow, be built, be imagined” -&gt “imposing”.

  43. Trond Engen says

    I should add that until the comma after “probably” this is essentially according to Bjorvand & Lindeman. From there and out it’s essentially worthless speculation.

  44. Further East, Kuznetsov (Russian), Kovalenko (Ukrainian), Kovalchuk (Belorussian), Kowalski (Polish), or simply Kuznets or Kowal are all Smiths. And they are tops in surnames.
    At one point, I had four Kovalenkos in my class at school. And later, at work, we had two Kuznetsovs with the same given name, patronymic and surname. Every payday there was trouble as cashiers tried to figure out who was who, as one was a highly paid senior editor and another a junior reporter.

  45. J.W. Brewer says

    My experience with objections to -person on the grounds that the syllable “son” was itself sexist was that they uniformly seemed to be jocular parodies of a style of feminism more humorless and obsessive than that of whoever was telling the joke (who might typically be a left-of-center self-identified feminist herself or himself). Although I suppose it is always hazardous to assume that a reductio-ad-absurdam joke about offensive language usage will never coincide with a sense of offense sincerely and seriously held by some sufficiently humorless activist out there.
    Of course, religious groups still differ with each other as to whether “parson” is or should be a male-only occupation.

  46. Where’s marie-lucie when we need her?
    Hee hee! She was at my house. So was Trond. By now she’ll be home, I think.

  47. Lucky you!

  48. David Marjanović says


    whereas Fabre is not particularly common in France

    This, too, is Occitan.
    See also: Cyrano de Donaldac, in whose presence the word “bill” must not be mentioned.

  49. Donald’s an American, so that’s “beak” (or perhaps “bec” in this case).

  50. Further East, Kuznetsov (Russian), Kovalenko (Ukrainian), Kovalchuk (Belorussian), Kowalski (Polish), or simply Kuznets or Kowal are all Smiths. And they are tops in surnames.
    So is Seppänen (Finland), or Sepp (Estonia), as well as derivatives of sepp – Kingsepp/Kingissepp (=shoemaker) and Rätsep (=taylor).

  51. Trond Engen says

    Scandinavian languages don’t do this. I used to think it had to do with the relative insignificance of urban settlement, but that’s equally true for Finnish and Estonian – and it’s not even true of the whole of Scandinavia. Also, until a few centuries ago, people were often listed with their occupation as a byname, here too. It may simply have to do with some rule for the late 19th / early 20th century transition to inherited surnames. But then the question is why the outcome was more or less the same in the three countries. (Iceland is a special case, since it went the other way and mounted obstacles against inherited surnames.)

  52. Trond Engen says

    It just struck me that the relation between bygg “barley” and byggva “build, make grow, make appear” is the same as the one between rís “rice” and risa “rise”. Except that the latter two are not etymologically related.

  53. Trond Engen says

    No need to envy the host. He’s probably still washing up. Envy me.

  54. David Marjanović says

    Oh, beak vs. bill is US vs. UK? I thought bill was more like, you know, duckbill… as in “duckbilled dinosaurs”…

  55. I don’t think of “bill” as British for “beak”. Like David, I think of it as the name for bird-mouths of the duckier kind.
    And surely “beak” as slang for nose is as British as it is American.

  56. Was it Dodger who used a Beak for a magistrate? In Oliver Twist? And the Bill or Old Bill is for London police.

  57. He’s probably still washing up.
    Funnily enough the next day the dishwasher broke.
    Trond gave me a magnificent present: De Indoeuropeiske Språkenes Historie, by Ole Wikander.

  58. If Americans didn’t use “bill”, they’d have to say “duck-beaked platypus”. As Ø says, bill and beak are not the same. The OED on bill:

    1. a.1.a The horny beak of certain birds, especially when slender, flattened, or weak.
       In Ornithology, beak is the general term applicable to all birds; in ordinary language beak is always used of birds of prey, and generally when striking or pecking is in question; beak and bill are both used of crows, finches, sparrows, perching birds and songsters generally, bill being however more frequent; bill is almost exclusively used of humming-birds, pigeons, waders, and web-footed birds.

    Americans also have bill to mean the peak of a cap, which either comes from its similarity to a bird’s bill or I’m a Dutchman.
    As well as a magistrate “the beak” is also a Victorian slang for a school’s headmaster (I think there are still schools that use it). As any reader of Billy Bunter knows, at some schools “beaks” is or was used by the boys for all the masters.
    On the (old) bill to mean the police, here are some related meanings for “bill” in the OED. I wonder if it comes from one of these (esp. 3.3):

    1.1 A weapon of war mentioned in OE. poetry, a kind of broadsword, a falchion. Obs. (Probably passing with modified shape into sense 2.)
       a 1000 Beowulf 4126 Æfter billes bíte.    c 1050 Ags. Gloss. in Wr.-Wülcker Voc. 376 Chalibem, bill.    1205 Lay. 1740 Þer wes bil ibeat ‹revsc› þer wes balu muchel.    [1867 Freeman Norm. Conq. (1876) I. v. 273 note, The bill here [in Death of Brihtnoth] spoken of was a sword and not an axe.]
    2.2 An obsolete military weapon used chiefly by infantry; varying in form from a simple concave blade with a long wooden handle, to a kind of concave axe with a spike at the back and its shaft terminating in a spear-head; a halberd.
       Distinct forms of bills seem to have been painted or varnished in different colours; hence the black bill and brown bills of the 16th and 17th centuries.
       c 1300 K. Alis. 1624 With longe billes‥They carve heore bones.    1465 Marg. Paston Lett. 518 II. 215 The tenauntes‥havyng rusty polexis and byllys.    1495 Act 11 Hen. VII, lxiv. Pream., Armours Defensives, as‥Bowes, Billes, Hauberts.    1593 Shakes. Rich. II, iii. ii. 118 Distaffe-Women manage rustie Bills.    1598 Barret Theor. Warres i. i. 2 Inveterate opinion‥touching blacke bils and bowes.    1603 Drayton Bar. Warres ii. xxxvii, Wer’t with the Speare, or Browne Bill, or the Pike.    1813 Scott Trierm. i. xiii, When the Gothic gateway frown’d, Glanced neither bill nor bow.    1834 J. R. Planché Brit. Costume. 33.
    b.2.b A similar weapon used by constables of the watch till late in the 18th cent. Also attrib.
       1589 Pappe w. Hatchet (1844) 28 All weapons from the taylors bodkin, to the watchmans browne bil.    1599 Shakes. Much Ado iii. iii. 44 Haue a care that your bills be not stolne.    1616 Fletcher Cust. Country ii. i. 9, 1. Off. He was still in quarrels, scorned us Peace-makes, And all our bill-authority.    1799 S. Freeman Town Off. 176 Every watchman carries a staff with a bill fastened thereon.
    3.3 Short for bill-man.
       1495 Hen. VII. in Ellis Orig. Lett. i. 11. I. 21 For‥an archer or bille on horsback viijd. by the day.    1513 Hen. VIII. in Strype Eccl. Mem. I. ii. App. i. 4 A hundred able men‥wherof threescore to be archers and forty bills on foot.    1532 G. Hervet Xenophon’s Househ. (1768) 35 Billes, and archers, the which folowe their capitaynes in good arraye.    1825 Scott Talism. x, A strong guard of bills and bows.

  59. Trond Engen says

    No. bile, Ger. Beil etc. means “broadaxe”.

  60. I grant that ducks have bills; I was overreacting.
    One of my favorite wisecracks: “The end of a lark is the beak.”

  61. Spanish surnames also show the importance of mills; Molina is the top occupational surname listed in the frequency tables of the Instituto Nacional de Estadística, in 30th position overall. Herrera and Herrero are also well-placed (57 and 72, respectively), and added together they would rank above Molina.
    I was surprised to find Guerrero in place 48; I can’t think of any equally-frequent analogues in other European languages.

  62. My bad. If I had read the article first, I would have noticed the high frequency of Kemp in English.

  63. Trond Engen says

    ‘Kemp’ = No. kjempe “giant” (&lt “big, strong guy” &lt “fighter” &lt the VLat. champion word). I didn’t know that.
    Kjempe is one of those words that (in some dialects) can be grammatically feminine even if it’s semantically masculine. I’ve been wondering if that’s because the Lat. ending -ion was identified with the Germanic common feminine ending *-jon-. (A more reliable one is bølle “bully”. “Han er ei bølle.”).

  64. John Friend says

    For all the people wondering about the lack of “Shoemakers” in English versus “Schumacher” in German, there are a number of reasons. 1) English language people adopted fixed surnames many centuries before German language people did. The English started adopting them between the years 1000-1300. Germans didn’t follow suit until the years 1600-1800. Before that, they were using patronymic surnames. 2) Because of this, most English surnames are centuries older than German surnames. 3) Shoemakers are leather workers. Another old-timey term for “leather worker” is “tanner”. 4) “Tanner” is a much more common surname in English than is “Shoemaker”. 5) The probable reason for this is that “shoemaker” as a specialized profession wasn’t nearly as prevalent in the years 1000-1300 as a more generalized “tanner” was. A “shoemaker” probably wouldn’t have been seen outside of London and maybe a few other larger cities. 6) Meanwhile, in future Germany, by the 1700s, there *were* specialized “Schumachers” both in the cities and in the countryside. 7) In addition to “Tanner” and “Shoemaker”, there is also the last name “Crispin”. The English, especially in the 1600-1700s, liked to use occupational patron saints instead of actual occupations as surnames. St. Crispin is the patron saint of shoemakers, so a lot of shoemakers ended up with the surname “Crispin”, especially, if they were foreigners from Germany, Holland, or Scandinavia (who didn’t yet use fixed surnames) coming to the UK or to North American in the 1600s and 1700s. This also cut down on the number of “Shoemaker”s. These same people who stayed in their home countries would have ended up as “Schumacher”s. 8) Lastly, there is also in the last name “Glover” in English which is another specialized leathermaking occupation. The surname is identical in German, but does not seem to be as prevalent there. So this one’s kind of a wash, but it does bring up an interesting aside: Crispin Glover’s name could accurately be interpreted as meaning “Leatherworker Leatherworker”.

  65. Very persuasive, and I love the aside!

  66. David Marjanović says

    Lastly, there is also in the last name “Glover” in English which is another specialized leathermaking occupation. The surname is identical in German, but does not seem to be as prevalent there.

    “Identical in German”? What do you mean? Is there a name Handschuhmacher? ~:-|

    BTW, don’t forget to add Schuster to Schu(h)macher.

  67. I do find a few dozen hits for “Herr Handschuhmacher” (but perhaps that just means “master glover”) and a dozen or so for “Frau Handschumacher”.

  68. David Marjanović says

    but perhaps that just means “master glover”

    Nope, that would be Handschuhmachermeister. (Not kidding.) Herr is “lord”, not “master”.

  69. Herr is “lord”, not “master”.
    Well, it’s “master” in the sense of “master and servant”, etc., not in the sence of “master artisan” or “guild master”.

  70. Ah. Well, as I write to people who address me as “Dr. Cowan”, neither doctor, master, nor bachelor am I, but plain John of Avalon (i.e. the island of the apple).

Speak Your Mind