This map shows the ordering of subject, object, and verb in a transitive clause, more specifically declarative clauses in which both the subject and object involve a noun (and not just a pronoun), as in the English sentence in (1).
(1)
[The dog]
chased
[the cat].
S
V
O
English is SVO (Subject-Verb-Object), because the subject the dog in (1) precedes the verb while the object the cat follows the verb.
There are six logically possible orders of the three elements S, O, and V, as shown in the feature-value table.
| Go to map | ||
|---|---|---|
| Value | Representation | |
 |
Subject-object-verb (SOV) | 564 |
 |
Subject-verb-object (SVO) | 488 |
 |
Verb-subject-object (VSO) | 95 |
 |
Verb-object-subject (VOS) | 25 |
 |
Object-verb-subject (OVS) | 11 |
 |
Object-subject-verb (OSV) | 4 |
 |
Lacking a dominant word order | 189 |
| Total: | 1376 | |
All six of these types are attested; examples of each type are given in (2).
(2)
a.
John
ga
tegami
o
yon-da.
John
subj
letter
obj
read-pst
S
O
V
‘John read the letter.’
b.
Mandarin (Li and Thompson 1981: 217)
Zhāngsān
shōudǎo-le
yi-fēng
xìn.
Zhangsan
receive-perf
one-clf
letter
S
V
O
‘Zhangsan received a letter.’
c.
Irish (Dillon and Ó Cróinín 1961: 166)
Léann
[na
sagairt]
[na
leabhair].
read.pres
the.pl
priest.pl
the.pl
book.pl
V
S
O
‘The priests are reading the books.’
d.
Nias (Austronesian; Sumatra, Indonesia; Brown 2001: 538)
i-rino
vakhe
ina-gu
3sg.realis-cook
abs.rice
mother-1sg.poss
V
O
S
‘My mother cooked rice.’
e.
Hixkaryana (Carib; Brazil; Derbyshire 1979: 87)
toto
y-ahosɨ-ye
kamara
man
3:3-grab-distant.pst
jaguar
O
V
S
‘The jaguar grabbed the man.’
f.
awad
kalapéé
hapʉ́h
jaguar
child
see.ind
O
S
V
‘The child sees the jaguar.’
Although all six of these orders are attested, the last two types, OVS and OSV, in which the object comes first, are rare.
The terms subject and object are used here in a rather informal semantic sense, to denote the more agent-like and more patient-like elements respectively. Their use here can be defined in terms of the notions S, A, and P, where the S is the single argument in an intransitive clause, the A is the more agent-like argument in a transitive clause, and the P is the more patient-like argument in a transitive clause. For the purposes of this map, then, the term subject is used for the A while the term object is used for the P. A language shown on the map as SOV could thus also be equally well and perhaps more accurately described as APV. Note that many linguists use the terms subject and object somewhat differently from this, and some linguists question the applicability of these terms to some languages, but these issues do not arise with the use of these terms here. For example, there is controversy surrounding the question of what ought to be considered the subject in Philippine languages, like Cebuano (cf. Schachter 1976). Cebuano has two common ways to express transitive clauses, one of which is illustrated in (3).
(3) Cebuano (Austronesian; own data)
gi-palit
[sa
babayi]
[ang
saging]
goal.foc-buy
nontop
woman
top
banana
‘The woman bought the bananas.’
While there is a question as to which of the two arguments in (3) should be considered a subject (or whether neither or both should), in both types of clauses the verb normally comes first, followed by the A, and then the P. Hence, by the use of subject and object assumed for this map, Cebuano is treated as a VSO language.
Note that while the position of the subject in intransitive clauses is generally the same as in transitive clauses, in some languages this is not the case. See Chapter 82.
Some languages can be assigned straightforwardly to one of the six types, because all orders other than one are either ungrammatical or used relatively infrequently and only in special pragmatic contexts. Such languages can be said to have rigid order. There are many other languages in which all six orders are grammatical. Such languages can be said to have flexible order. Flexible order languages are sometimes described as having “free” word order, though this is misleading, since there are often pragmatic factors governing the choice of word order.
We can further distinguish two subtypes of languages with flexible word order. In some languages with flexible order, there is one order which is most common and which can be described as the dominant order. In other flexible order languages, the flexibility is greater and there is no one order that is the dominant order in terms of frequency of usage or pragmatic neutrality. Flexible order languages in which one order is dominant are shown on the map according to that dominant order – in other words, the map does not distinguish rigid order languages from flexible order languages with a dominant order. Flexible order languages lacking a dominant order are shown on the map as “lacking a dominant word order”. Russian is an example of a language with flexible word order in which SVO order can be considered dominant, so Russian is shown on the map as SVO. See “Determining Dominant Word Order”.
There are a number of different subtypes of languages lacking a dominant order which are not distinguished on the map. In some languages with highly flexible word order, all or most orders of subject, object, and verb will be possible and common. Nunggubuyu (Gunwinyguan; northern Australia) is an example of such a language (Heath 1984: 507-513; 1986). But some languages lack a dominant order only because just the subject or just the object exhibits flexibility with respect to the verb. For example, Syrian Arabic allows both SVO and VSO orders and there does not seem to be a reason (at least on the basis of the description by Cowell 1964: 407, 411) to consider one of them dominant. However, only these two orders are common and the order of verb and object is relatively inflexible.
A third subtype of language lacking a dominant order consists of languages in which different word orders occur but the choice is syntactically determined. For example, in German and Dutch, the dominant order is SVO in main clauses lacking an auxiliary and SOV in subordinate clauses and clauses containing an auxiliary (see below for examples). Because this results in both orders being common, neither order is considered dominant here and these two languages are shown on the map as lacking a dominant word order. In general, if the word order varies according to whether there is an auxiliary verb, the language is shown on the map as lacking a dominant order. Another language whose word order depends both on whether there is an auxiliary and whether the clause is a main clause is Dinka (Nilotic; Sudan): like German, the order is SVO in main clauses without an auxiliary, SAuxOV in main clauses with an auxiliary, but it is VSO in subordinate clauses without an auxiliary and AuxSOV in subordinate clauses with an auxiliary (Nebel 1948: 9, 25, 42, 75, 82).
Where languages differ in their order between main clauses and subordinate clauses, the order in main clauses is used to classify them on this map. For example, Quileute (Chimakuan; Washington State) is VSO in main clauses and SVO in subordinate clauses (Andrade 1933: 278), and is shown on the map as VSO. In some languages, word order is more fixed in subordinate clauses. For example, in Miya (Chadic; Nigeria), while both SVO and VOS are found in main clauses, only VOS order is found in adverbial subordinate clauses and relative clauses (Schuh 1998: 281, 291); because both SVO and VOS are common in main clauses, Miya is shown on the map as lacking a dominant order.
The most frequent of the six orders is SOV and it is widely distributed across the globe. Perhaps the most striking region in which SOV predominates is an area covering most of Asia, except in Southeast Asia and the Middle East. It is also overwhelmingly the dominant order in New Guinea, most of the exceptions being along the north coast. It is the most common order among languages in Australia which have a dominant order at all, although even in languages in which SOV is dominant, the order is generally flexible. It is clearly the dominant order in North America outside of the Pacific Northwest and Mesoamerica.
The map shows three areas where SVO order predominates: (i) an area covering much of sub-Saharan Africa, though with a scattering of SOV and VSO languages; (ii) an area extending from China and southeast Asia south into the Austronesian languages of Indonesia and the western Pacific; and (iii) Europe and around the Mediterranean. SVO order is not common outside these areas.
VSO order is scattered around various parts of the world, in eastern Africa (among various Eastern Sudanic languages), in North Africa (Berber), in the western extremes of Europe (Celtic), in and around the Philippines, among Polynesian languages of the Pacific, in Mesoamerica, and in the Pacific Northwest. VOS order is also scattered around the globe, though there are no attested instances on the mainland of Africa or Eurasia.
There are nine OVS languages on the map, six of which are spoken in South America, five in the Amazon basin, and one (Selknam) in Tierra del Fuego. There are only four OSV languages shown: Warao in Venezuela, Nadëb in Brazil, Wik Ngathana in northeastern Australia, and Tobati in West Papua, Indonesia. Languages without a dominant order are especially common in North America and Australia, and to a lesser extent in South America. The scattering of this type partly reflects the fact that this is not a homogeneous type, since it mixes languages with highly flexible order with languages which have more rigid order but where there are two dominant orders. The former type, languages with highly flexible order, is most common in North America and Australia and relatively uncommon in Africa, Europe, Asia, New Guinea and among Austronesian languages.
The map in figure 1 shows the distribution of word order in various languages of the past, though the times at which these languages were spoken vary from 4500 years ago to 1000 years ago. The map illustrates the fact that SVO, now a common order in Europe and around the Mediterranean, was less common in the past: on the one hand, there were SOV languages like Latin and Etruscan in western Europe; on the other hand, there were many VSO languages in what is now the Middle East, represented both by Semitic languages and by Egyptian.
Map 81B shows languages where it is not possible to identify a single dominant order but in which it is possible to identify two orders, neither of which is dominant relative to the other, but which can be said to be dominant relative to other orders.
| Go to map | ||
|---|---|---|
| Value | Representation | |
 |
SOV or SVO | 29 |
 |
VSO or VOS | 14 |
 |
SVO or VSO | 13 |
 |
SVO or VOS | 8 |
 |
SOV or OVS | 3 |
| Total: | 67 | |
Type 1 represents languages which are SOV/SVO, i.e. languages in which the orders SOV and SVO are common relative to other orders, but where neither order is dominant relative to the other. An example of a language of this type is German, as illustrated in (4).
(4) German
a.
Der
Lehrer
trink-t
das
Wasser.
def
teacher
drink-3sg
def
water
‘The teacher is drinking the water.’
b.
Der
Lehrer
ha-t
das
Wasser
getrunken.
def
teacher
have-3sg
def
water
drink:past.ptcpl
‘The teacher has drunk the water.’
Note that at the level of grammar, German is simply verb-second, placing the inflected verb second, and when the inflected verb is an auxiliary verb, the main verb occurs at the end of the clause. Thus when the inflected verb is the main verb, it is possible to get OVS order, or VSO order if an adverb or prepositional phrase precedes the verb, but SVO is apparently more common than these. Type 1 is the most common of the types on this map; this most likely is related to the fact that SOV and SVO are by far the two most common types among the six possible orders of subject, object, and verb on Map 81A.
Type 2 represents languages which are SVO/VSO. The fact that there are half as many languages of this type compared to SOV/SVO languages while there are more than five times as many SOV as VSO languages means that the ratio of SVO/VSO to VSO is higher than the ratio of SOV/SVO to SOV. In other words, the frequency of Type 2 is greater than we might expect. The high number of Type 2 languages relates to Universal 6 of Greenberg (1963): “All languages with dominant VSO order have SVO as an alternative or as the only alternative basic order”. Although there may be exceptions to Greenberg's universal, the fact that VSO languages generally allow SVO as an alternate word order makes it unsurprising that we would find many SVO/VSO languages. A number of the Type 2 languages occur in genealogical groups that also contain VSO languages, such as Berber, Semitic, and various subgroups of Austronesian.
Type 3 languages represent languages which are VSO/VOS. This is the second-most frequent type on Map 81B and given the fact that neither of the two types it is based on are among the two more frequent orders SOV and SVO, this type is considerably more frequent than one might otherwise expect.
Type 4 represents languages which are SVO/VOS, analogous to Type 2, except with VOS instead of VSO. Their high frequency despite the relative rarity of VOS languages is presumably analogous to the factors discussed above with respect to Type 2.
Type 5 represents languages which are SOV/OVS. The three languages in the sample of this type are Wichita (Caddoan; Oklahoma), Apalaí (Cariban; Brazil), and Macushi “Determining Dominant Word Order”. Note that a number of OVS languages are fairly rigidly OV, but allow SOV as a not infrequent alternate order, so that in some cases there may be a question as to whether a language should be classified as OVS or SOV/OVS.
Some languages are described as being verb-final with the order of subject and object pragmatically determined, suggesting that they are of a type SOV/OSV not shown on this map, such as Anamuxra (Madang; Papua New Guinea) (Ingram 2001) and Mekeo (Oceanic; Papua New Guinea) (Jones 1998a). While there may be some languages of this type, it has been my experience that when one counts the frequency of SOV and OSV order in texts for languages which are so described, SOV usually outnumbers OSV order by well over 2 to 1, and thus the language counts as SOV by the criteria assumed in this chapter (see “Determining Dominant Word Order”). In the absence of such text counts, such languages are shown on Maps 82A and 83A as SV and OV, but are not shown on this map, it being considered unknown whether they are SOV or SOV/OSV. I have yet to find a language where text counts show SOV and OSV occurring with comparable frequency and hence I have yet to find a language that I would classify as SOV/OSV.
Among the languages shown on Map 81B, there are some languages where the alternation between the two orders is apparently pragmatically governed while there are others where it is grammatically governed. German is a clear example of the latter sort. In fact, the languages where the alternation is grammatically governed are generally of Type 1. A number of languages (Nuer (Ethiopia and Sudan), Dinka (Sudan) both Nilotic languages, and Ngiti (Lendu; Democratic Republic of Congo)) employ the SOV order when there is an auxiliary, although in Nuer the auxiliary element is nonverbal. The Moru-Ma'di languages (represented here by Moru (Sudan), Avokaya (Democratic Republic of Congo and Sudan), Logoti (Democratic Republic of Congo), Lugbara (Democratic Republic of Congo and Uganda), and Ma'di (Sudan and Uganda)) have two constructions, a perfective/past construction and an imperfective/nonpast one, the former being SVO, the latter SOV, and when an auxiliary is employed in the SOV construction, the order is SAuxOV.
While the feature shown on this map is perhaps the single most frequently cited typological feature of languages, it is now recognized that it represents a clause type that does not occur especially frequently in spoken language; it is more common that at least one of the two arguments of a transitive clause will be pronominal, and in many languages pronominal subjects are expressed by verbal affixes. It is argued by Dryer (1997) that a more useful typology is one based on two more basic features, whether the language is OV or VO and whether it is SV or VS; these are shown on the next two maps, in Chapters 82 and 83. In addition, as noted above, the order in transitive clauses is not always the same as the order in intransitive clauses. The feature shown on this map is also important in that many other features are predictable from it, at least statistically. Most of these features correlate more specifically with the order of object and verb (Greenberg 1963, Dryer 1992; see Chapters 95, 96, and 97). For a few features, SVO languages exhibit properties intermediate between those of SOV languages and those of verb-initial languages, though in general they are more similar to verb-initial languages (Dryer 1991).