Naamkunde. Jaargang 4

[p. 26]

The Place-Name Element -hurst (-horst).

As a place-name element hurst (horst, hyrst, host, etc.) is attested over a wide area in both Germany and the Netherlands. In northwestern Germany it ranges over Southwest Schleswig, Holstein, North and Lower Saxony, Westfalia, and is sparsely attested from East Friesland. In the Netherlands it is attested from the provinces of North Holland, South Holland, Utrecht, Gelderland, Overijsel, Drente, Brabant and Limburg. It occurs frequently as a field-name in the netherlandic part of Belgium, from West Flanders to Limburg.⁽¹⁾ In West Friesland harst (< OFr *herst < *hyrst) is only infrequently attested as a place-name formant, and it does not occur as an appellative.⁽²⁾ Given weak attestation of hurst as a place-name

(1) See further, M. Schönfeld, Veldnamen in Nederland (Mededelingen der Koninklijke Nederlandse Akademie van Wetenschappen, Afd. Letterkunde N.R., Deel 12, No. 1, 1950), pp. 33-4, particularly p. 33, fn. 188 with literature. Schönfeld states that horst-names are very well known ‘besides in the “Zuidelijke Nederlanden” (= Northern Belgium), in northern Germany, partly by netherlandic colonization in that area’ (page 33). For the northern part of Belgium horst-names are listed by K. de Flou, Woordenboek der Toponymie van Westelijk Vlaanderen (1914-1938, 18 vols) for Aarsele: 1432 de hurst; 1846 den heust and for Dentergem: 1415 de huuerst, 1474 de hurst, 1474 de hurst-meersch. For Brabant J. Lindemans mentions horst or hurst as synonyms of hage ‘bos van laagstammig hout’ in Beersel: 1593 hurst of haghe; in Opwijk: 1440, 1725 de horst, de hursten; 1750 d'heurst (Handelingen Com. Top. Dial. 5 (1931), 222, note 2). For the province Antwerpen horst/hurst occurs in the unpublished monographies of the Instituut voor Naamkunde (Louvain): in Wilrijk: 1639 den horst, 1704 de horsten; in Brecht: 1834 den horst; in Geel: 1679 horst, 1856 heust; in Herentals: 1696 de horst, 1720 de hoorst; in Mol: 1562 den horst, 1681 de horsten; in Weelde: 1643 de horst and so on In the province of Limburg we find in Zonhoven: 1433 die hoerst, 1440 dy horst, 1601. die horsten, 1470 horst bampt, 1462 horsland, 1479 horststraet; in Kaulille: 16th cent. horstbeem.
H.J. Moerman, Nederlandse plaatsnamen (Leiden, 1956), pp. 105-7, states that horst is a Saxon word and notes that horsten in the Beekbergen Wood, Veluwe, were wooded eminences above marshland. The oldest attestations of Netherlandic horst-names are from Overijsel and Gelderland: Mikulunhorst (Ov. 933-966), Binkhorst (Ov. 933-966) and Bochursti (Geld. 806).

(2) See J.J. Kalma, Fryske Plaknammen (Leeuwarden, 1949), I, pp. 39-40. The place-name files of the Fryske Akademy (Leeuwarden) list: Harst (Westergeast), Oppe Harst (Blije, Dearsum), Harstbuorren (Dearsum) in addition to Brande Harst.

[p. 27]

[p. 28]

formant in East Friesland and non-occurrence as an appellative or place-name element in North Frisian and West Flanders, we conclude that Frisian place-names in -harst probably derive from Saxon influence.

In Middle Dutch the appellative is by no means as well attested as in Middle Low German, and it apparently passed out of the language in the 17th century.⁽³⁾ In his Dictionarium teutonico-latinum (1574) Kiliaan lists horscht [sic] as: ‘virgultum, silva humiles tantum frutices proferens frutetum, senticetum’. In view of the non-occurrence of hurst as appellative and place-name formant in West Flemish and its concentration as a place-name formant in the Saxon areas of the Netherlands, we conclude that place-names in -hurst (-horst) appear to have spread to the Netherlands from Saxon Germany.

In western middle and southern Germany hurst is attested as an appellative with the phonological and semantic realizations indicated on Map 1. In these areas of Germany hurst is infrequently attested as a field-name, generally designating poor or fallow land, land overgrown with thorns and bushes. In Switzerland hurst is found in field-names in Bern, Glarus, Aargau, Wallis and Basel.⁽⁴⁾ In the Rhineland field-names in -hurst are found in the Ruhr. In Old High German hurst is attested as a gloss (dumus) and in three place-names.⁽⁵⁾

In northeastern Germany hurst is found as both appellative and place-name formant in Mecklenburg, around the Vistula Delta, in northern West Prussia, and infrequently in Southern Hannover and Braunschweig, as well as in Altmark. Mitzka⁽⁶⁾ states that hurst-names are attested in Mecklenburg from the 13th century, in Brandenburg and the Vistula Delta from the 14th century and concludes, in which Schönfeld⁽⁷⁾ concurs, that they possibly reflect Dutch influence. Bathe⁽⁸⁾ cites hurst-names from Altmark and infrequently

(3) E. Verwijs and J. Verdam, Middelnederlandsch Woordenboek (The Hague, 1885- 1928), III, col. 605.

(4) Schweizerisches Idiotikon (Frauenfeld, 1881-), II, 1640-41.

(5) E.G. Graff, Althochdeutscher Sprachschatz (Berlin, 1835-1846), IV, 1042. - E. Steinmeyer and E. Sievers, Die althochdeutschen Glossen (Berlin, 1897-1922), II, 713.15.

(6) W. Mitzka, Nordseeküste und ostdeutsche Ortsnamen, Zeitschrift für Ortsnamenforschung 9 (1933), pp. 3-5.

(7) M. Schönfeld, Het onderzoek der plaatsnamen van Nederlandse oorsprong in de laatste jaren, NGN 9 (1934), p. 7.

(8) M. Bathe, Die Herkunft der Siedler in den Landen Jerichow, erschlossen aus der Laut-, Wort- und Flurnamen-Geographie (Halle, 1931), pp. 104-105.

[p. 29]

from Börde and Fläming and concludes that they replaced dunke-names c. 1300 in the designation of elevated areas in moorland. Hurst-names in northeastern Germany are necessarily younger than in northwestern Germany: they were introduced to the East from the West as the result of combined Low German-Dutch colonization beginning in the 12th century.

Hurst (hyrst, herst, hirst) is well attested from Old English and Middle English as both appellative and place-name formant. As a masc. a-stem it is glossed ‘hillock, height, wooded eminence, wood’. Hurst-names are generally restricted to Saxon England and concentrated in Hampshire, Surrey, Sussex and Kent, being most frequent in the latter two shires with no fewer than forty names from Sussex clone.⁽⁹⁾ They rarely occur in East Anglia, where they represent a younger level of names than in Saxon England.⁽¹⁰⁾ The paucity of hurst-names in East Anglia may coincide with the corresponding paucity of harst-names in Frisian, for Frisian influence was perhaps most prominent in this portion of England.

An appellative *haurst or the like is not attested from Gothic, nor is hurst (horst, hyrst, etc.) attested as an appellative or place-name formant from the older stages of the Scandinavian languages. In the modern period horst is attested only from Danish and represents a Modern loan from the turn of the century. It is not attested from the dialects in the collections at Institut for Dansk Dialektforskning, nor is it attested from the numerous Danish dialect monographs and dictionaries.⁽¹¹⁾ We therefore conclude that hurst is uniquely West Germanic.

In the older etymological dictionaries it is generally assumed that the Norwegian and Swedish appellative and place-name formant rust (ryst, röst) is the result of metathesis of West Germanic hurst.⁽¹²⁾

(9) See R.G. Roberts, The Place-Names of Sussex (Cambridge, 1914) and J. Wallenberg, The Place-Names of Kent (Uppsala, 1934).

(10) See H. Jellinghaus, Die westfälischen Ortsnamen (Kiel & Leipzig, 1896), 2nd ed., p. 53 ff. and Idem, Englische und niederdeutsche Ortsnamen, Anglia 20 (1898), pp. 295-6 for a comparative Low German-English corpus.

(11) For the dialect information I am indebted to Ella Jensen and Chr. Lisse of Institut for Dansk Dialektforskning. Prof. John Kousgard Sorensen has informed me (per litteras) that Borghorst (Dänischwohld, South Schleswig) is the only example of a hurst-name in a formerly Danish area and, hence, the only example of a hurst-name in Scandinavia. However, Borghorst is primarily a German settlement.

(12) H.S. Falk and Alf Torp, Norwegisch-Dänisches etymologisches Wöorterbuch (Oslo & Bergen, 1960), 2nd ed. II pp. 923-4. - Elof Hellquist Svensk etymologisk ordbok (Lund 1939), II, col. 854 b. For the Nw. appellative rust/röst, see Ivar Aasen, Norsk Ordbog (Christiania, 1873). - H. Ross, Norsk Ordbog (Christiania, 1895-1913). The Sw. appellative attestations were kindly forewarded to me from SOAd (Svensk ortnamnsarkivet, dialektordssamling) by Dr. Allan Rostvik.

[p. 30]

In the older period the appellative occurs once in Old Icelandic in the compound garđsrust, ‘collapsed wall’, in Sturlunga Saga. The Norwegian appellative rust fem. has a general distribution and is glossed ‘copse, mountain ridge, bank of stones from a collapsed wall’. The Swedish appellative röst fem. (and neut. in Särna, northern Dalacarlia) is only attested from formerly Norwegian areas: Jämtland, Häarjedal and Särna. It is glossed ‘copse, mountain ridge, low-lying swampy area’, and the appellative rust/röst is thus apparently restricted to West Norse.

The distribution of Scandinavian rust/röst-names is shown on Map 2.⁽¹³⁾ The highest concentration of rust-names in Norway is in Oppland, and the oldest Norwegian attestations are both from Lom Parish, Oppland: Staurust ( rússt), Staurrust DN 5:173 (1355) and Aukrust (krúst), Aukrust DN 15:24 (1367). Norwegian rust-names are both compounded and uncompounded, and in compounds rust is invariably the second member, while the first is never a personal name. As is the case for German and English hurst-names, Norwegian rust-names are most frequently compounded with tree names. The same holds true for Swedish names in -röst, and, with but few exceptions, Swedish röst-names are restricted to western Sweden in formerly Norwegian areas from Jämtland to Bohuslän.

Previously, we noted derivation of rust < hurst, an etymology rejected by de Vries,⁽¹⁴⁾ who regards rust as a verbal abstract and derives it from Germanic *rusti-, an -s- extension of IE *reu-/*rū̆-,

(13) This is a class of nature names which, as far as I know, has been totally neglected in the literature; their distribution is mapped here for the first time. In his introduction to Norske Gaardnavne (Oslo, 1897), p. 72. O. Rygh merely states that, though the appellative is not known from the older stages of Scandinavian, there are many old names in -rust. Attestations for Norway were excerpted from Norske Gaardnavne (Christiana-Oslo, 1897-1936), I-XIX. The Swedish material was excerpted from SOA by Dr. Allan Rostvik. The exceptional eastern Swedish attestations are: Landröste (Ösma Parish, Södertörn, Södermanland), Slättrösten (Kristberg Parish, Östergötland) and Storrösten (Los Parish, Hälsingland).

(14) So also J. Pokorny, Indogermanisches etymologisches Wörterbuch (Bern, 1959), p. 870. Note the archaic suffixation in -ti- with zero grade root vocalism and feminine gender, cf. E. Benveniste, Origines de la formation des noms en indo-européen (Paris, 1962), 3rd ed. pp. 81-4.

[p. 31]

[p. 32]

‘to dig, heap up’. Cf. OE rēost, ‘wooden socket for the plow blade’, and note OS, OHG riostra. This formation is attested only from Germanic and Baltic and represents a further lexical isogloss bounding the two dialects. Note Lithuanian raũsti, ‘to heap up earth, dig’, Latvian raust, ‘to poke, heap up goods’. The ‘copse’ gloss for Scandinavian rust/röst may be explained as secondary, deriving from reference to growth on mounds of boulders, gravel, etc. In view of the discontinguous distribution of hurst and rust, the reflection of i-umlaut in Swedish röst and the divergent semantic fields of rust and hurst, the derivation advanced by de Vries is the most plausible. If so, then West Germanic hurst and North Germanic rust are unrelated.

The area with the highest concentration of hurst-names in Germany is Westfalia, located in the heart of the northwestern German st-name area.⁽¹⁵⁾ Westfalia has several hundred hurst-names, while Lower Saxony has c. ninety and East Friesland attests only a few such names. Westfalian hurst-names are attested from c. 850, and both area of concentration and relative chronology bear witness to the resettlement of the Saxons south of the Elbe after the Anglo-Saxon landnám.⁽¹⁶⁾ Generally speaking, hurst-names in northwestern Germany devote a sandy, dry area overgrown with shrubs or bramble elevated above a moor or swamp. For example, from the traditionally conservative Lüneburger Heide, Kück⁽¹⁷⁾ gives horst/host as ‘an elevated clearing on which only shrub grows, presumably a recently cleared area, possibly by burning’. Hurst fem. is not attested in the Heliand and occurs only in the lesser texts,⁽¹⁸⁾ but it is frequently attested as both appellative and place-name element in Middle Low German.⁽¹⁹⁾ Old English hurst-names are compounded in the same manner as Old Saxon hurst-names, i.e. with animal, bird and geographical lexemes. Both Continental and Insular hurst-names are rarely compounded with personal names, cf. Kentish Froecinghyrst and Westfalian Freckenhorst (Fricconhurst, 851).⁽²⁰⁾

(15) Hans Kuhn Vor- und frühgermanische Ortsnamen in Norddeutschland und den Niederlanden Westfälische Forschungen 12 (1959) 5-44.

(16) Jellinghaus Die westfälischen Ortsnamen p. 53 ff.

(17) E. Kück Lüneburger Wörterbuch (Neumüster 1942-1967) I col. 748.

(18) J.H. Gallée Vorstudien zu einem altniederdeutschen Wörterbuch (Leiden,1903), p. 155.

(19) K. Schiller and A. Lübben, Mittelniederdeutsches Wörterbuch (Bremen, 1875- 1881), II, cols. 304b-305a. - A. Lübben, Zu Reineke Vos, Germania (Pfeiffer) 8 (1863), 370-3.

(20) Jellinghaus, Englische und niederdeutsche Ortsnamen, pp. 295-6.

[p. 33]

The concentration of hurst-names in Saxon England, as well as the equivalent manner of composition of Continental and Insular hurst-names, points to Continental Saxon origin and sets a terminus ante quem at c. 450 for their formation. We conclude that hurst-names antedate the Anglo-Saxon landnám, are generally Saxon and bear witness to the resettlement of the Continental Saxons south of the Elbe after the landnám. On the Continent they probably spread from a central area of concentration in Westfalia. In northeastern Germany, Holland and Schleswig-Holstein they represent a younger level of names.

Both appellative and place-name element are generally semantically equivalent within the domain of such names, and the two basic recurrent semantic features of hurst are: 1) elevation and 2) undergrowth, i.e. a thicket of secondary growth on an elevated area, usually a sandy mound.

Hurst is generally derived from le *kert-/*kṛt-, ‘turn, twist together, compact’, + -s-ti-.⁽²¹⁾ Alternatively, Osthoff⁽²²⁾ derives it from IE *ker-/*kor-, ‘grow’, reflected in Latin Cerēs, crebēr, Oscan kerrí, etc. However, these etymologies must be rejected on semantic grounds alone when we consider OE hyrstan (frigere), a secondary denominative of hyrst, OHG geharstit (frixam), harsti (frixura).⁽²³⁾ Note the col

(21) So Falk-Torp, Norwegisch-Dänisches etymologisches Wörterbuch, p. 924. - Hans Krahe, Über st-Bildungen in den germanischen Sprachen, PBB 71 (1949), p. 243. - Fr. Kluge, Etymologisches Wörterbuch der deutschen Sprache (Berlin, 1967), 20th ed., p. 317. - W. Meid, Über s in Verbindung mit t-haltigen Suffixen, besonders im Germanischen, IF 69 (1964), pp. 229-30. Pokorny (Indogermanisches etymologisches Wörterbuch, p. 584) tentatively derives hurst < *kṛt-s-ti-, but subsequently (p. 633) considers this etymology less probable and proposes derivation from *kṷṛs-to-, which Meid, loc. cit., rightfully rejects. Derivation from *kṛt-s-ti- would yield a Gmc. *hurs(s) in accordance with Bartholom ae's Law, but there is good reason to doubt the absolute invariability of the prescribed assimilation. Note OE hloest, OHG hlast vd. OIc. hlass < *klət-to-. Derivation from *kṛt-s-ti- cannot be rejected on phonological grounds. Note further the etymology recently proposed by Jost Trier, Horst und Stock, in Gedenkschrift für William Foerste (Cologne and Vienna, 1970 = Niederdeutsche Studien, 18), pp. 100-108, from IE State I *kṷer- /II *kṛ-, ‘cut, hew, break, pull up’. Trier fails to account for umlauted OE forms, but even more significant is his failure to note the connection between secondary growth, general topographical restriction to elevated areas and the association of horst-names with burning, which is hardly coincidental, see infra.

(22) H. Osthoff, Etymologische Parerga (Leipzig, 1901), I, pp. 57-8, cf. Pokorny, Indogermanisches etymologisches Wörterbuch, p. 577.

(23) Steinmeyer-Sievers, Die althochdeutschen Glossen, I, 416.26, II, 222.69. - J. Bosworth and T.N. Tobler, Anglo-Saxon Dictionary (Oxford, 1882-1898), s.v. hyrst.

[p. 34]

location of hurst and brennen in Gottfried's Tristan (c. 1210): daz er hürste vil von grime abe brande und ūz der erden sluoc 9002. In this connection recall Kück's comment that Lüneburger horst might designate an area cleared by burning and note West Frisian Brande Harst. Thus hurst must have covered a third basic semantic field: ‘fire, burn, heat’. We propose a derivation from IE *ker-/*kṛ-, ‘burn, glow, heat’, reflected in Lithuanian kùrti, ‘to heat’, kūréntis, ‘be heated’, Latin cremō, Umbrian krematra, Gothic hauri, ‘coals’ (John 18:18, Romans 12:20), Old Icelandic hyrr < *huriR. We then posit derivation from Germanic base forms: *hurist-/*hursti-.⁽²⁴⁾ Trees as the object burned are indicated both by Gothic hauri and Latin carbō, where -b- is of uncertain origin.

The use of fire is fundamental in primitive slash-and-burn agricultures, and it is with reference to this practice and the distribution of the majority of older hurst-names on sandy mounds or ridges that the other semantic components (elevation and secondary growth) receive clarification. First, secondary growth is the natural result of areas cleared by burning. Second, the concentration of hurst-names in northwestern Germany and Saxon England with a general topographical restriction to mounds or ridges elevated above unarable marshland is an indication of where such a clearing process could best take place. Slash-and-burn agriculture is necessarily impermanent when practiced in sandy soils and normally employed by hunting and fishing peoples. These facts, together with the postulation that hurst-names reflect this practice, fit together very well with Caesar's often cited observations on Germanic customs in Book IV of his

(24) In Gmc. -st was rarely added to roots ending in vowels or -r, though frequently suffixed to stem vowels in primary and secondary derivations, as OHG dionō-st:dionō-n, see W. Meid, ‘Über s’, pp. 254-5. Reasonable alternatives to account for umlauted hyrst/herst are therefore *hur-i-st/*hur-st-i. Subsequent suffixal contamination with -ti-/-tu-, cf. Goth. luftus vs. OE lyft, might be advanced to account for unumlauted vs. umlauted forms, see K. von Bahder, Die Verbalabstracta in den germanischen Sprachen (Halle, 1880), pp. 94-5. However, suffixal contamination is not strictly motivated, cf. OE fȳst and OHG fūst < *funksti-. Note OS Hursti, Hursteromarka and MDu. -hursti > -horst c. 1200. On the other hand, syncopation of -i- after short stem syllables is not altogether unknown in WGmc., e.g. better, best, cf. A. Walde, Die germanischen Auslautgesetze (Halle, 1900), pp. 8-9. Cf. OE đyrs masc. i-stem and OIc. Þurs. OHG duris/thuris are ghost forms; the attested form is turs in Notker. Nevertheless, derivation from *hursti- > hyrst/hurst seems the more plausible alternative. OS hurst > MLG horst merely reflects regular development of u > o before r, see A. Lasch, Mittelniederdeutsche Grammatik (Halle, 1914), art. 61.

[p. 35]

Gallic Wars: Sed est nihil privati ac separati agri apud eos; neque licet remanere longius anno in uno loco, causa incolendi; neque vivunt multum frumento, sed maximam partem lacte atque pecore. Finally, then, we conclude that hurst-names may bear witness to the early practice of slash-and-burn agriculture among the West Germanic peoples and thereby represent an old level of especially Saxon Rodungsnamen, just as names in -tveit/-tved represent specifically Scandinavian settlement names of the same genre.

Harvard University.

T.L. Markey.

Postscriptum

After submission of the above article, Prof. H. Draye has kindly called my attention to an article by Fritz Langenbeck, Die tung-und hurst-Namen im Oberrheinland, Alemannisches Jahrbuch (1958) 51-108. Langenbeck is not, as I have been, concerned with the etymology of hurst and its wider distribution as both appellative and place-name element. The central purpose of Langenbeck's paper is to lend further support to the entirely probable thesis that hurst-names in Middle Baden reflect Saxon colonization initiated in the Carolingian period, for as a productive place-name formant hurst is notably exclusively Saxon in origin.

Samenvatting

De verspreiding van hurst, horst, hyrst, host als simplex of als tweede bestanddeel bij boom-, vogel-, natuurnamen (niet bij persoonsnamen) wijst op een concentratie in N.W. Duitsland (inzonderheid in Westfalen), midden in het voorhistorisch -st -gebied van H. Kuhn (zie kaart nr. 1). Van dit ‘Saksisch’ kerngebied schijnt hurst/horst te zijn uitgezwermd naar het Nederlands taalgebied, naar N.O. Duitsland en naar Saksisch England (Sussex en Kent), sporadisch slechts naar Friesland. Hurst/horst-toponiemen duiden meestal een verhevenheid aan, met secundaire begroeiing na voorafgaande oorbaarmaking door vuur. S. leidt hurst/horst af van IE. *ker/*kṛ, Germ. *hurist, *hursti (‘brand, gloed, hitte’) (vgl. o.m. Lat. cremo) en verwerpt een etymologische samenhang van Wgm. *hurist, *hursti met Skandinavische toponiemen rust, ryst, röst), zie kaart nr. 2.

[H.D.]