European hydrotoponymy (IV): tug of war between Balto-Slavic and West Uralic

In his recent paper on Late Proto-Indo-European migrations, when citing Udolph to support his model, Frederik Kortlandt failed to mention that the Old European hydrotoponymy in northern Central-East Europe evolved into Baltic and Slavic layers, and both take part in some Northern European (i.e. Germanic – Balto-Slavic) commonalities.

Proto-Slavic

From Expansion slavischer Stämme aus namenkundlicher und bodenkundlicher sicht, by Udolph, Onomastica (2016), translated into English (emphasis mine):

NOTE. An archived version is available here. The DOI references for Onomastica do not work.

(…) there is a clear center of Slavic names in the area north of the Carpathians. Among them are root words of the Slavic languages such as reka / rzeka, potok u. a. m.

Even more important than this mapping is the question of how the dispersion of ancient Slavic names happened. What is meant by ancient Slavic names? I elaborated on this in this journal years ago (Udolph, 1997):

(1)Ancient suffixes that are no longer productive today.

This clearly includes Slavic *-(j)ava as in Vir-ava, Vod-ava, Il-ava, Glin-iawa, Breg-ava, Ljut-ava, Mor-ava, Orl-java among others. It has clear links to the ancient common Indo-European language (Lupawa, Morava-March-Moravia, Orava, Widawa). They have a center north of the Carpathians.

ava-slavic

(2) Unproductive appellatives (water words), which have disappeared from the language, are certain witnesses of ancient Slavic settlements. A nice example of this is Ukr. bahno, Pol. bagno ‘swamp, bog, morass’ etc. The word has long been missing in South Slavic, although it appears in South Slavic names, but only in very specific areas (see Udolph, 1979, pp. 324-336).

(3) Names that go back to different sound shifts. [Examples:]

  • (…) the Slavic clan around Old Sorbian brna ‘feces, earth’, Bulgarian OCS brьnije ‘feces, loam’, OCS brъna ‘feces’, Slovenian brn, ‘river mud’, etc. is solved with the inclusion of onomastic materials (Udolph, 1979, p. 499-514). (…) Toponymic mapping shows important details.
  • bryn-slavic
    Karte 4. brъn < *brŭn und bryn- < *brūn- in slavischen Namen
  • (…)We also have an ablauting *krŭn-:*krūn- in front of us. Map 5 shows the distribution of both variants in Slavic names.
  • The next case is quite similar. It concerns Russ. appellative grjaz’ ‘dirt, feces, mud’, (…) for which an Old Slavic form *gręz exists. Slavic also knows the ablauting variant *grǫz.

    These maps (see Map 6, p. 222) show that a homeland of Slavic tribes can only be inferred north of the Carpathians.

    (4) Place-names formed by Slavic suffixes of Pre-Slavic nature, i.e. derived from Old European hydronyms.

    (a) The largest river in Poland, the Wisła, German Vistula, bears a clearly Pre-Slavic name, no matter how one explains it (Babik, 2001, pp. 311-315; Bijak, 2013, p. 34, Udolph, 1990 , Pp. 303-311).

    (b) With the same suffix are formed Sanok, place on the southwest of Przemyśl; Sanoka, a no longer known waters name, 1448 as fluvium Szanoka, near the place Sanoka and with a diminutive suffix -ok- a tributary of the Sanok, which is called Sanoczek (for details see Udolph, 1990, pp. 264-270; Rymut / Majtan, 1998, p. 222). The San also has a single-language name, but that does not change anything about the right etymology. The suffix variant -očь also includes Liwocz and Liwoczka, river names near Cracow; also a mountain range of the Beskydy is mentioned at Długosz as Lywocz.

    According to the opinion of the “Słownik prasłowiański” (Sławski (red.), 1974, p. 92), the suffix -ok- represents a Proto-Slavic archaism. It appears, for example, in sъvědokъ, snubokъ, vidokъ, edok, igrok, inok among others, but its antiquity also shows, among other things, that it started at archaic athematic tribes.

    east-slavic-language-expansion
    Mapping of older and younger East Slavic place-names and translation into settlement evolution.

    Slavonic Urheimat

    If we apply this to the loess distribution in western Ukraine and south-eastern Poland, it is very noticeable that the center of the Old Slavic place names lies in the area where loess dispersal is gradually “frayed out”, i.e. for example, in the area west of Kiev between Krakow in the west and Winnycja and Moldavia in the east. In short, the distribution of good soils coincides with ancient Slavic names. If that is correct, we can expect a homeland in the Pre-Carpathian region, or better, a core landscape of Slavic settlement.

    The existence of Pre-Slavic Indo-European place names and water names whose structure indicates that they originated from an Indo-European basis, but then also developed Slavic peculiarities, can now – as stated above – only be understood to mean that the language group that we call today Slavic emerged in a century-long process from an Indo-European dialectal area.

    Loess areas between Poland and Ukraine. Image from Jary et al. (2018).

    From a genetic point of view, the scarce data published to date show a clear shift of central-east populations from more Corded Ware-like groups in the EBA towards more BBC-derived ancestry in the common era, to the point where ancient DNA samples from East Germany, Poland and Lithuania evolve from clustering between Corded Ware and Sub-Neolithic peoples to clustering close to Bell Beaker-derived groups, such as West Germanic peoples, Tollense samples, etc. (see below)

    Furthermore, sampled Early Slavs show bottlenecks under “Dinaric” I2a-L621 and central-eastern E1b-V13, which – in combination with the known phylogeography of Únětice and Urnfield – is compatible with its late expansion from a central-east European Slavonic homeland, such as the Pomeranian culture, in turn likely derived from Lusatian culture groups.

    This doesn’t preclude a more immediate expansion of Common Slavic in Antiquity closer to the northern Carpathians, which is also supported by the available Early Slavic sampling, apart from samples from the Avar and Hungarian polities.

    pca-balto-slavic-iron-age
    Likely Baltic (yellow-green) and Slavic (orange) groups ca. 500 AD on, with Finnic (cyan) and Mordvinic (blue) groups roughly divided through hydrotoponymy line ca. 1000 AD Top Left: Late Iron Age cultures. Top right: PCA of groups from the Iron Age to the Middle Ages. Y-DNA haplogroups during the Germanic migrations (Bottom left) and during the Middle Ages (Bottom right). Notice a majority non-R1a lineages among sampled Early Slavs. See full maps and PCAs.

    Proto-Baltic / Proto-Slavic

    Northern European hydronymy

    From Alteuropäische Hydronymie und urslavische Gewässernamen, by Udolph, Onomastica (1997), translated into English (emphasis mine):

    NOTE. An HTML version is available at Jurgen Udolph’s personal site.

    Because of the already striking similarities as the well-known “-m-case”, the number-words for ‘1000’, ’11’ and ’12’ and so on, J. Grimm had already assumed a close relationship between Germanic and Baltic and Slavic. (…)

    In my own search, I approached this trinity from the nomenclature side. In doing so, I noticed some name groups that can speak for a certain common context:

    1.* bhelgh-, *bholgh-.

    Map 10, p. 64, shows that a root * bhelgh- occurs in the name material of a region from which later Germanic, Baltic and Slavic originated. The Balkans play no role in this.

    bholgh-germanic-balto-slavic

    2. *dhelbh-, *dholbh-, *dhl̥bh-

    The proof of the three ablauting * dhelbh, * dholbh, * dhl̥bh- within a limited area shows the close relationship that this root has with the Indo-European basis. Again it is significant in which area the names meet (…)

    dhelbh-germanic-balto-slavic

    3. An Indo-European root extension *per-s- with the meaning ‘spray, splash, dust, drop’ is detectable in several languages (…). From a Baltic-Slavic-Germanic peculiarity cannot therefore be spoken from the toponymic point of view. The picture changes, however, if one includes the derived water names.

    4. The root extension *pel-t-, *pol-t-, *pl̥-t- of a tribe widely spread in the Indo-European languages around *pel-, pol- ‘pour, flow, etc.’, whose reflexes are found Armenian through Baltic and Slavic to the Celtic area, is found in the Baltic toponymy, cf. Latv. palts, palte ‘puddle, pool’.

    trzciniec-riesenbecher-culture
    The dynamics of stylistic changes of the form of the “Trzciniec pot” in the lowland regions of Central Europe, and spreading routes of the Trzciniec package in Central Europe. A good proxy for contacts through the Northern European Plain during the Early Bronze Age. Modified from Czebreszuk (1998).

    Early Balto-Finnic

    In order to properly delimit (geographically and chonologically) the Proto-Baltic and Proto-Slavic expansions, it is necessary to understand where the late Balto-Finnic homeland was located during the Bronze Age. The following are excerpts from the comprehensive hydrotoponymic study by Pauli Rahkonen (2013):

    In any case, Finnic probably had its origin somewhere around the Gulf of Finland. Names of large and central rivers such as Vuoksi (< Finnic vuo ‘stream’) and Neva (< Finnic neva ‘marsh, river’) must be very old and might represent Proto-Finnic hydronyms. In the southern coastal area of Finland, the names Kymi and Nietoo < *Niet|oja (id. later Porvoonjoki) may also be of Finnic origin and derive from, respectively, kymi ‘stream’ (see SSA I s.v. *kymi; see however SPK s.v. Kemijärvi; Rahkonen 2013: 24) and nieto(s) ‘heap of snow’ (SSA II s.v. nietos), in hydronyms probably ‘high (snowy?) banks of a river’. Mustion|joki is clearly a Finnish name < *must|oja ‘black river’. The river name Vantaa remains somewhat obscure, although Nissilä (see SPK s.v. Vantaanjoki) has derived it from the Finnic word vana ‘water route’. In western Finland the names of large rivers, such as Aura and Eura, are supposedly of Germanic origin (Koivulehto 1987).

    In Estonia the names of many of the most important rivers might be of Finnic origin: e.g. Ema|jõgi Est. ema ‘mother’ [Tartu district] (?? cf. the Lake Piiga|ndi < Est. piiga ‘maiden’), Pärnu [Pärnu district] < Est. pärn ‘linden’, Valge|jõgi [Loksa district] < Est. valge ‘white’, Must|jõgi [Võru district] < Est. must ‘black’. It is possible that Emajogi and especially Piigandi are the result of later folk etymologizing of a name with some unknown origin. However, as a naming motif there exist in Finland numerous toponyms with the stems Finnic *emä (e.g. 3 Emäjoki), *neit(V)- ‘maiden’ (e.g. Neitijärvi, Neittävänjoki, Neittävänjärvi) and Saami stems that can be derived from Proto Saami *nejte̮ ‘id’ (GT2000; NA).

    finnic-toponyms
    The historical southern boundary of Finnic hydronyms, excluding hydronyms produced by the Karelian refugees of the 17th century.

    These seemingly very old names of relatively large rivers in southern Finland, modern Leningrad oblast and Estonia support the hypothesis that Proto-Finnic was spoken for a long time on both sides of the Gulf of Finland and it thus basically corresponds to the hypothesis of Terho Itkonen (see below). In the Novgorod, Tver or Vologda oblasts of Russia, Finnic names for large rivers cannot be found (Rahkonen 2011: 229). For this reason, it is likely that the Late Proto-Finnic homeland was the area around the Gulf of Finland.

    Beyond the southeastern boundary of the modern or historically known Finnic-speaking area, there exists a toponymic layer belonging to the supposedly non-Finnic Novgorodian Čudes (see Rahkonen 2011). In theory it is possible that Proto-Finnic and Proto-Čudian separated from each other at an early stage or it is even possible that Proto-Čudian was identical with Proto-Finnic. However, this cannot be proven, because there is not enough material available describing what Novgorodian Čudic was like exactly.

    finno-saamic-mordvin
    Yakhr-, -khra, yedr-, -dra and yer-/yar, -er(o), -or(o) names of lakes in Central and North Russia and the possible boundary of the proto-language words *jäkra/ä and *järka/ä. Rahkonen (2013)

    A summary of the data is then:

    • The Daugava River and the Gulf of Livonia formed the most stable south-western Balto-Finnic border (up until ca. 1000 AD): the Daugava shows a likely Indo-European etymology, while some of its tributaries are best explained as derived from Uralic.
    • The first layer of “Early Baltic” loans in Early Balto-Finnic are of a non-attested Baltic dialect closest to Proto-Balto-Slavic (read more about this early layer).
    • The latest samples of the Trzciniec culture (or derived Iron Age group) from its easternmost group in Turlojiškė (ca. 1000-800 BC?) show a western shift towards Bell Beaker, although they show a majority of hg. R1a-Z280; while the earliest sample from Gustorzyn (ca. 1900 BC), likely from Trzciniec/Iwno, from the westernmost area of the culture, shows a Corded Ware-like ancestry (and hg. R1a-Z280, likely S24902+) among a BA sampling from Poland clearly derived from Bell Beaker groups.

    One can therefore infer that the expansion of the Trzciniec culture – as the earliest expansion of central-west European peoples into the Baltic after the Bell Beaker period – represented either the whole disintegrating Balto-Slavic community, or at least an Early Baltic-speaking community expanding from the West Baltic area to the east.

    The similarity of Early Slavs and the Trzciniec outlier with the Czech BA cluster, formed by samples from Bohemia (ca. 2200–1700 BC), and the varied haplogroups found among Early Slavs – reminiscent of the variability of the Unetice/Urnfield sampling – may help tentatively connect the early Proto-Slavic homeland more strongly with a Proto-Lusatian community immediately to the south-west of the Iwno/Proto-Trzciniec core.

    pca-late-bronze-age-balto-slavic-finnic
    Top Left:Likely Baltic, Slavic, and Balto-Finnic-speaking territories (asynchronous), overlaid over Late Bronze Age cultures. Balto-Slavic in green: West(-East?) Baltic (B1), unattested early Baltic (B2), and Slavic (S). Late Balto-Finnic (F) in cyan. In red, Tollense and Turlojiškė sampling. Dashed black line: Balto-Slavic/West Uralic hydrotoponymy border until ca. 1000 AD. Top right: PCA of groups from the Early Bronze Age to the Late Bronze Age. Marked are Iwno/Pre-Trzciniec of Gustorzyn (see below), Late Trzciniec/Iron Age samples from Turlojiškė, and in dashed line approximate extent of Tollense cluster; Y-DNA haplogroups during the Late Bronze Age (Bottom left) and during the Early Iron Age (Bottom right). Notice a majority non-R1a lineages among sampled Early Slavs. See full maps and PCAs.

    Proto-Balto-Slavic homeland

    Disconnected western border: Germanic

    The common Balto-Slavic – Germanic community must necessarily be traced back to the West Baltic. From Udolph’s Namenkundliche Studien zum Germanenproblem, de Gruyter (1994), translated from German (emphasis mine):

    My work [Namenkundliche Studien zum Germanenproblem] has shown how strong the Germanic toponymy is related to the East, less to Slavic, much more to Baltic. It confirms the recent thesis by W.P. Schmid on the special relationship Germanic and Baltic, according to which “the formation of the typical Germanic linguistic characteristics…must have taken place in the neighborhood of Baltic“.

    If one starts from a Germanic core area whose eastern boundary is to be set on the middle Elbe between the Erzgebirge and Altmark, there are little more than 400 km. to the undoubtedly Baltic settlement area east of the Vistula. Stretching the Baltic area westwards over the Vistula (as far as the much-cited Persante), the distance is reduced to less than 300 km. Assuming further that Indo-European tribes between the developing Germanic and the Baltic groups represent the connection between the two language groups, so can one understand well the special relationship proposed by W.P. Schmid between Germanic and Baltic. In an earlier period shared Slavic evidently the same similarities (Baltic-Slavic-Germanic peculiarities).

    balto-slavic-balto-finnic-homeland
    Top: Palaeo-Germanic (G2, blue area), Proto-Balto-Slavic/Pre-Baltic (PBSL, green area) and Early Proto-Balto-Finnic (PBF, cyan area) homelands superimposed over Early Bronze Age cultures. Persante hydronym and Gustorzyn ancient DNA sample location marked. Y-DNA haplogroups during the Early Bronze Age (Bottom left) and during the Middle Bronze Age (Bottom right). Notice a mix of R1b-L151 samples from the west and the process of integration of R1a-Z645 lineages from the the north-east. See full maps and PCAs.

    Substrate and immediate eastern border: Early Balto-Finnic

    While Balto-Finnic shows a late Balto-Slavic adstrate, Balto-Slavic has a Balto-Finnic(-like) substrate, also found later in Baltic and Slavic, which implies that Balto-Slavic (and later Baltic and Slavic) replaced the language of peoples who spoke Balto-Finnic(-like) languages, influencing at the same time the language of neighbouring peoples, who still spoke Balto-Finnic (or were directly connected to the Balto-Finnic community).

    For more on this relative chronology in Balto-Slavic – Balto-Finnic contacts, see e.g. the recent posts on Kallio (2003), Olander (2019), or a summary of this substrate.

    While Rahkonen (2013) entertains Parpola’s theory of a West-Uralic-speaking Netted Ware area (ca. 1900-500 BC), due to the Uralic-like hydrotoponymy of its territory, he also supports Itkonen’s idea of the ancient presence of almost exclusively Balto-Finnic place and river names in the Eastern Baltic and the Gulf of Finland since at least the Corded Ware period, due to the lack of Indo-European layers there:

    NOTE. This idea was also recently repeated by Kallio (2015), who can’t find a non-Uralic layer of hydrotoponymy in Balto-Finnic-speaking areas.

    It should be observed that the territory between the historical Finnic and Mordvin-speaking areas matches quite well with the area of the so-called Textile Ceramics [circa 1900–800 BC] (cf. Parpola 2012: 288). The culture of Textile Ceramics could function as a bridge between these two extreme points. Languages that were spoken later in this vast territory between Finland–Estonia and Mordovia seem to derive from Western Uralic (WU) as well. I have called those languages Meryan-Muroma, Eastern and Western Čudian and an unknown “x” language spoken in inland Finland, Karelia and the Lake Region of the Russian North (Rahkonen 2011; 241; 2012a: 19–27; 2013: 5– 43). This might mean that the territory of the Early Textile Ceramics reflects to some extent the area of late Western Uralic.

    The archaeologically problematic area is Estonia, Livonia and Coastal Finland – the area traditionally assumed to have been populated by the late Proto-Finns. The Textile Ceramics culture was absent there. It is very difficult to believe that the Textile Ware population in inland Finland migrated or was even the main factor bringing the Pre- or Early Proto-Finnic language to Estonia or Livonia. There are no archaeological or toponymic signs of it. Therefore, I am forced to believe that Textile Ceramics did not bring Uralic-speaking people to those regions. This makes it possible, but not absolutely proven, to assume that some type of Uralic language was spoken in the region of the Gulf of Finland already before Textile Ceramics spread to the northwest (circa 1900 BC).

    corded-ware-west-uralic
    Top Left: Corded Ware culture expansion. Top right: PCA of Corded Ware and Sub-Neolithic groups. Y-DNA haplogroups during the Corded Ware expansion (Bottom left) and during the subsequent Bell Beaker expansion (Bottom right). Notice the rapid population replacement of typical Corded Ware R1a-Z645 lineages by expanding Bell Beakers of hg. R1b-L23 in central-east Europe, while they show continuity in the described ancestral Fennoscandian West-Uralic-speaking territory. See full maps and PCAs.

    The Corded Ware population in Finland is thought to have been NW Indo-European by many scholars (e.g. Koivulehto 2006: 154–155; Carpelan & Parpola 2001: 84). At least, it is probable that the Corded Ware culture was brought to Finland by waves of migration, because the representatives of the former Late Comb Ceramics partially lived at the same time side by side with the Corded Ware population. However, it is possible that the immigrants were a population that spoke Proto-Uralic, who had adopted the Corded Ware culture from their Indo-European neighbors, possibly from the population of the Fatjanovo culture, e.g. in the Valdai region. This was suggested by Terho Itkonen (1997: 251) as well. In that case the population of the Typical and Late Comb Ceramics may have spoken some Paleo European language (see Saarikivi 2004a). In the Early Bronze Age, the Baltic Pre-Finnic language that I have suggested must have been very close to late WU and therefore no substantial linguistic differences existed between the Baltic Pre-Finns and the population of Textile Ceramics in inland Finland. I admit that this model is difficult to prove, but I have presented it primarily in order to offer new models of thinking.16 At least, there is no archaeological or linguistic reason against this idea.

    This dubitative attribution of Proto-Uralic to the expansion of Corded Ware groups in eastern Europe, which is what hydrotoponymic data suggests in combination with archaeology, has to be understood as a consequence of how striking Rahkonen finds the results of his research, despite Itkonen’s previous proposal, in the context of an overwhelming majority of Indo-Europeanists who, until very recently, simplistically associated Corded Ware with the Indo-European expansion.

    Conclusion

    Even Kortlandt accepts at this point the identification of expanding East Bell Beakers from the Carpathian Basin as those who left the Alteuropäische layer reaching up to the Baltic. However, he identified Udolph’s data solely with West Indo-European, forgetting to mention the commonly agreed upon western Proto-Balto-Slavic homeland, most likely because it contradicts two of his main tenets:

    1. that Balto-Slavic split from a hypothetical Indo-Slavonic (i.e. Satem) group expanding from the east; and
    2. that laryngeals can be reconstructed for Balto-Slavic – unlike for North-West Indo-European.
    old-european-asian-hydro-toponymy
    Indo-European hydrotoponymy in Europe and the Middle East (scarce Central Asian data). Baltic data compensated, statistical method RBF: intermediate regions devoid of Indo-European toponyms are inferred to have them; it compensates thus e.g. for the scarce Indo-European hydrotoponyms in Poland by assuming ‘soft’ continuity from West Germany to the Baltic.

    A hypothetic “Pre-Indo-Slavonic” laryngeal Indo-European layer reaching Fennoscandia and the Forest Zone with Corded Ware is fully at odds with all known data:

    • in comparative grammar, since the one feature that characterizes Graeco-Aryan is precisely its set of innovations relative to Northern Indo-European, which presupposes a longer contact (and further laryngeal loss) once Tocharian and North-West Indo-European had separated – hence probably represented by Palaeo-BalkanCatacomb-Poltavka contacts once Afanasevo and Yamna settlers from the Carpathian Basin / East Bell Beakers had become isolated;
    • in hydrotoponymy, because of the prehistoric linguistic areas that can be inferred from (1) the distribution of Old European hydrotoponymy; (2) Udolph’s work on Germanic and the likely non-Indo-European substrate in Scandinavia and land contacts with Balto-Finnic; (3) from the Northern European traits in the Northern European Plain; or (4) from the decreasing proportion of Indo-European place and river names from central Europe towards the east and north.
    • NOTE. An alternative explanation of Old European/Balto-Slavic layers, e.g. by a ‘Centum’ Temematic – even if one obviates the general academic rejection to Holzer’s proposal – couldn’t account for the absolute lack of an ancestral layer of Indo-European hydrotoponymy in North-Eastern Europe (i.e. the longest-lasting Corded Ware territory), in sharp contrast with Western Europe, South-Eastern Europe, and South Asia. All of that contradicts an Eastern Indo-European community, even without a need to recall that the oldest hydrotoponymic layers common to Fennoscandia and the Forest Zone are of Uralic nature.

    • in archaeology, because cultural expansions of the Eastern European Early Bronze Age province since the Bell Beaker period (viz. Mierzanowice, Trzciniec, Lusatian, Pomeranian, West Baltic Culture of Cairns) suggest once and again west-east movements, most (if not all) of which – based on the presence of Indo-European speakers during the common era – were likely associated with Indo-European-speaking communities replacing or displacing previous ones.
    • in palaeogenomics, because of the late and different association of Corded Ware ancestry and haplogroups among Balto-Slavic and Indo-Iranian communities, in turn corresponding to the different satemization processes found in both dialects, which may have actually been related to the Uralic substrate that is found in both (read more on Uralic influences on Balto-Slavic and on Indo-Iranian).

    On the other hand, a careful combination of Uralic and Indo-European comparative grammar, hydrotoponymic data, and population genomics fits perfectly well Itkonen’s and Rahkonen’s association of Corded Ware in Eastern Europe with Uralic languages, as well as the traditional mainstream view of Uralic before Indo-European in Fennoscandia and in the Forest Zone, as I explained in a recent post about genetic continuity in the East Baltic area.

    Population genomics is not the main reason to reject the Indo-European Corded Ware theory – or any other prehistoric ethnolinguistic identification, for that matter. It can’t be. This new field offers just the occasional confirmation of a well-founded theory or, alternatively, another nail in the coffin of fringe theories that were actually never that likely, but seemed impossible to fully dismiss on purely theoretical grounds.

    The problem with Corded Ware was that we couldn’t see how unlikely its association with Indo-European languages was until we had ancient DNA to corroborate archaeological models, because few (if any) Indo-Europeanists really cared about the linguistic prehistory of eastern and northern Europe, or about Uralic languages in general (contrary to the general trend among Uralicists to be well-versed in Indo-European studies). Now they will.

    Related