Summary
$e objective of our article is to present the selected results of the research which was conducted 
for the purpose of our master’s thesis. We focused on the transfer of the functions of the present 
perfect into the domain of the preterite in informal British and American English. We put 
together a British and an American corpus and analysed the differences and similarities between 
British and American English in this transfer. We examined some factors that may influence 
this transfer either in American or British English, or in both varieties. $e major factors will be 
presented in this article. $e results show that the differences between both varieties mostly occur 
in the frequency of this transfer. Moreover, in American English there are more significant factors 
than in British English. Nevertheless, we can observe that the general tendency is the same in 
both varieties. $is fact may indicate that this phenomenon, which was first noticed in informal 
American English, is spreading to informal British English. 
Key words: the present perfect, preterite, transfer, British English, American English.
 Povzetek
Namen našega članka je predstaviti izbrane rezultate raziskave, ki je bila narejena za namen naše 
magistrske naloge. Osredinjali smo se na prehod funkcij dovršnega sedanjika v domeno navadnega 
preteklika v neformalni britanski in ameriški angleščini. Sestavili smo britanski in ameriški korpus 
besedil in analizirali razlike in podobnosti med britansko in ameriško angleščino pri tem prehodu. 
Pregledali smo dejavnike, ki lahko vplivajo na ta prehod bodisi v ameriški ali britanski angleščini 
bodisi v obeh variantah. Glavne dejavnike bome predstavili v tem članku. Rezultati kažejo, da se 
razlika med variantama pojavlja predvsem v pogostnosti tega prehoda. Nadalje je v ameriški angleščini 
mnogo več statistično pomenljivih dejavnikov kot v britanski angleščini. Vseeno pa opažamo, da se 
pojavlja enaka težnja v obeh variantah. T o dejstvo lahko nakazuje, da se ta pojav, ki so ga najprej opazili 
v neformalni ameriški angleščini, začenja širiti tudi na neformalno britansko angleščino.
Ključne besede: dovršni sedanjik, navadni preteklik, prehod, britanska angleščina, ameriška 
angleščina.
 DOI: 10.4312/elope.7.1.27-46

Our article aims at presenting the selected results of the research which was conducted for the 
purpose of our master’s thesis. We focused on the transfer of the functions of the present perfect 
into the domain of the preterite in informal British and American English. $e following sub-
section presents the research done by other linguists dealing with the same phenomenon. It also 
focuses on their varying views on this transfer. 
 $e review of the general literature shows that grammarians do not pay a great deal of attention 
to the transfer of the functions of the present perfect into the domain of the preterite in either 
AmE or in BrE. Comrie (1976, 53–4) mentions it only in one of his footnotes. We can find some 
information in the grammars of Quirk et al. (1985, 194–5), Leech (1987, 38–9), Leech and 
Svartvik (1994, 70) and Greenbaum (1996, 272) and the encyclopaedias of McArthur (1992, 
39) and Crystal (1995, 225, 311). One claim that they have in common is that the tendency of 
replacing the present perfect by the preterite occurs above all in the resultative and experiential 
meaning of the present perfect (i.e. in indefinite meaning). $is tendency extends also to non-
American varieties (Crystal 1995, 225) and is considered substandard (McArthur 1992, 45). 
McArthur (1992, 39) also claims that the transfer in AmE mostly occurs in situations that lead 
up to the present moment (i.e. in continuative meaning). 
A detailed treatment of this transfer can only be observed in the linguists who explicitly deal with 
this phenomenon (e.g. Vanneck 1958; Peterson 1970; Dušková 1976; Marshall 1979; Richards 
1979; Sheen 1984, 1994; Elsness 1997;). Some of them mainly focus on AmE (e.g. Vanneck 
1958; Peterson 1970; Richards 1979; Marshall 1982; Sheen 1984;), whereas some others do 
make mention of an important grammatical difference between AmE and BrE in the use of the 
present perfect and preterite (e.g. Dušková 1976; Sheen 1994; Elsness 1997).
As was discerned by Vanneck in 1958, there is a tendency towards replacing the present perfect 
by the ‘colloquial’
1
 preterite only in spoken AmE, whereas there is no such tendency in written 
BrE and AmE and also spoken BrE. Vanneck (1958, 238) emphasizes that “the colloquial 
preterite is a genuine preterite, not a perfect with the auxiliary slurred or omitted”. He focused 
on the everyday speech of educated native-speakers of American English. He finds that very few 
speakers of AmE are aware that their use of the present perfect differs from that of BrE speakers. 
He does not mention which meanings are replaceable; however it is apparent that he has the 
resultative and experiential present perfect (i.e. indefinite meaning) in mind. 
 ‘Colloquial’ preterite: a term used for the preterite that behaves as a free variant of the present perfect.

On the other hand, Marshall (1979) believes that the distinction between formal and informal 
AmE is far superior to that between the spoken and written AmE. $e object of her study was 
spoken AmE and she included people of different social classes and ages into her research. Her 
research shows that speakers use the colloquial preterite to a large extent; however they themselves 
judged it to be less correct and acceptable than the present perfect.
 
Peterson (1970) deals exclusively with AmE. She analysed ten American plays and divided them 
into constructions according to the presence of adverbials. She finds that the transfer can occur in 
both constructions. $ough Peterson does not explicitly mention the meanings where this transfer 
is possible, it is evident that the author has the resultative and experiential meaning (i.e. indefinite 
meaning) in mind. She does not define her concept of continuative meaning, and we therefore 
suppose that she assigns this meaning to the constructions with the adverbials for..., since..., in... 
and all. She claims that they are used with the present perfect because their use with the preterite 
would change the meaning. Peterson (1970, 5) believes that “(....) the choice between the present 
perfect and simple past seems to involve style rather than any rule of grammar”. 
$is transfer is treated in a similar way by Peterson and Dušková (1976), the only difference 
being that the latter includes BrE besides AmE into her research. $e object of Dušková’s study 
is both the traditional and the colloquial preterite. She finds that the preterite is more frequent in 
AmE than in BrE, which could indicate that in AmE it takes over some functions of the present 
perfect. Dušková focuses above all on temporal modifiers (i.e. temporal adverbials), an initiative 
for her article being that of Vanneck (1958), which, as she claims, needs certain corrections. Her 
research shows that the use of the colloquial preterite is not only typical of spoken AmE, but 
also of spoken BrE. In the latter the preterite is used above all with just, never, ever and always 
although this phenomenon is more frequent in AmE (Dušková 1976, 59). Dušková (1976, 54) 
adopts the traditional approach in the treatment of the differences between these two tenses. 
She focuses on temporal modifiers. As she claims, the use of a certain tense with a temporal 
modifier becomes redundant if the part of the meaning, usually expressed by the tense alone, is 
expressed semantically. When the semantics of temporal modifiers is sufficient to indicate the 
temporal reference intended, the use of the present perfect is redundant and consequently it can 
be replaced by the preterite. $is occurs more frequently in AmE than in BrE. When the semantics 
of temporal modifiers is not sufficient, then the present perfect’s and the preterite’s functions are 
kept well apart in both varieties. It is apparent that the transfer frequently occurs in resultative and 
experiential meaning (i.e. indefinite meaning). However, she also traces the continuative meaning, 
which is only rarely used with the preterite in the constructions with since..., in... and for.... As to 
be considered is that the preterite with continuative meaning can be used with temporal modifiers 
as all my life and all these years. It follows that the meaning does not have an important role in the 
transfer. According to her claim (1976, 58), the fact that the period that leads up to the present is 
established semantically (i.e. by temporal modifiers) is of crucial importance in the transfer. 
Sheen’s research from 1984 focuses on AmE, whereas his research from 1994 focuses on the 
difference between BrE and AmE in this transfer. He finds that the transfer also occurs in BrE. 
$e use of the preterite with the adverbials ever and never in BrE has long existed according 

to him. However, his research of BrE reveals that this transfer extends to other adverbials. $e 
difference between BrE and AmE lies above all in the frequency of this transfer, this transfer 
being of a less frequent occurrence in BrE. Sheen presumes that BrE may be under increasing 
influence of AmE. According to him, the finding that the colloquial preterite is more frequent 
in the spoken than in written language is not correct since there is a frequent occurrence of the 
colloquial preterite in the sports pages of newspapers. 
Richards (1979) claims that the resultative meaning of the present perfect is the only one that 
can be expressed by both the present perfect and the colloquial preterite. According to him, all 
other meanings remain in the domain of the present perfect. It is worth mentioning that he does 
not recognize that the colloquial preterite can take over the functions of the present perfect of 
experience – this being a disadvantage of his analysis, which has been stressed by Sheen (1984, 
377), who finds that the colloquial preterite may also convey other non-resultative meanings. 
Besides, Sheen (1984, 1994) concentrates on the constraints of this transfer. He is interested in 
under what circumstances the transfer is not possible. $e essential role in the interpretation of this 
transfer, as he claims, may be given to Current Relevance $eory (henceforth: CRT). However, he 
is convinced that CRT is not sufficient in this interpretation; therefore the combination of CRT 
and the theory of redundancy must be taken into consideration. In his model, current relevance 
(henceforth: CR), a basic feature of the present perfect, becomes redundant when it is established 
either by the context (i.e. linguistic context) or the situation (i.e. extra-linguistic context). $e 
temporal adverbials are a typical example of the linguistic context. $ey are supposed to create 
redundancy of conveying CR by the present perfect so that they themselves establish a period 
that leads up to the present and in this way also convey CR. $is approach is very close to 
Dušková’s (1976), who claims that the period must be semantically established. 
Furthermore, Sheen (1984, 379) finds the importance of “situational redundancy” (i.e. the extra-
linguistic context) in the transfer. In his conclusion, he says that the colloquial preterite takes 
over the functions of the present perfect both in spoken and also informal written language, 
albeit under certain constraints. According to his findings (382–3), only the present perfect is 
used: (1) “In verbal forms expressing a state which still exists at the time of speaking.”; (2) “In 
verbal forms used with since, for, and now and in subordinate clauses containing when, after, 
before, until, and as soon as; and (3) “In verbal forms in situations where the use thereof creates a 
state in the light of which future events must be viewed.” His article (1994, 45) brings additional 
finding: (4) in the present perfect constructions without temporal adverbials if the replacement 
with the colloquial preterite would denote a past situation without CR.
Hence it follows that the phenomenon of “free variation”, as it is called by Sheen, is possible in 
indefinite and continuative meaning with an exception in constructions with for..., since... and 
now. Other situations, as in constructions with always and ever that can lead up to the moment 
of speaking, can be used with the colloquial preterite. 
In his extensive research composed of three parts, Elsness (1997, 229), among other issues, studies 
the difference between the present perfect and the preterite in BrE and AmE. He finds that the 

present perfect is more frequent in BrE than in AmE. In his elicitation test, which was designed 
to find the difference between the BrE and AmE use of the colloquial preterite, he observes that 
the preterite is more acceptable in AmE, whereas in BrE it is the present perfect. $e difference 
mostly occurs in borderline cases, which are, as he claims, the situations that do not lead up to 
the deictic zero-point and are not tied to the anchor that denotes a clearly distinct definite time 
in the past. Borderline cases have anchors that are mostly temporal adverbials which express 
unclearly distinct recent indefinite time where the separation from the deictic zero-point is not 
clearly indicated. $ese adverbials are just, recently, always, never, ever, already and yet. Elsness 
does not base his interpretation of this transfer on CRT , but on different temporal schemata of 
these two tenses. He finds that the difference between both varieties occurs in the borderline 
cases where speakers of BrE and AmE take a different choice.
Elsness (1997, 229−36) studies a great many factors to find the difference between the present 
perfect and the preterite in BrE and AmE. Moreover, Elsness finds that CR does not play a 
significant role in distinguishing the differences between the varieties, although CR may gain a 
certain effect when there are no other factors that could affect the tense choice in a construction 
or in the wider linguistic or extra-linguistic context. As he finds, the important factor may be 
the structure of the verbal phrase and its immediate surrounding. $is can especially be applied 
to the spoken language where the distinction between these two tenses is not entirely clear-cut. 
His elicitation test demonstrates that in instances when it would be unusual to use the auxiliary 
in its uncontracted form (i.e. in affirmative constructions) and when the forms of the preterite 
and past participle are identical, speakers of BrE and AmE frequently choose the colloquial 
preterite. On the other hand, speakers of both varieties choose to use the present perfect when 
the uncontracted auxiliary is acceptable even in informal speech (i.e. in negative and interrogative 
constructions) and when the forms of the preterite and past participle are different.
 $e research was done for the purpose of our master’s thesis, where we focused on the transfer 
of the functions of the present perfect into the domain of the colloquial preterite. We put 
together a British and an American corpus, excerpting the material from plays. Each corpus 
consists of 360 constructions with the present perfect or the colloquial preterite. We sought out 
the first 60 constructions of these verbal constructions in each play. $e original intent was to 
analyse 6 British and 6 American plays (i.e. 360 constructions in each corpus). However, some 
of the chosen American plays were too short; thus some additional were necessary to obtain 
60 constructions. $e present research focuses on the transfer of the functions of the present 
perfect into the domain of the colloquial preterite in informal BrE and AmE and is limited to 
finite perfect and preterite constructions. Non-finite perfect constructions are excluded from the 
coding. Modal auxiliaries are also not the subject of this research. We put the corpus material 
into the computer programme Microsoft Excel 7, then determined the categories (factors) and 
values of all categories. We used the tools of the Microsoft Excel 7 for counting the constructions, 
percentage calculations and shaping of the tables and figures. We chose the Pearson’s χ
2
−test 
(95% significance level) (Ferligoj 1995, 160−6) to check the statistical significance of the factors 

that were supposed to affect the frequency of the transfer. When the factor was statistically 
significant at 99% significance level, this information was also noted.
 T o achieve our basic aim we sought and analysed:
(1) the differences and similarities between BrE and AmE in the transfer of the present perfect 
into the domain of the colloquial preterite, and
(2) the factors that affect the transfer in order to be able to compare the varieties. 
Above all, it is worth emphasizing that discussing adverbials is not the subject of this article. As a 
consequence they are not dealt with in the following sub-sections. Our research definitely includes 
both those constructions with and without adverbials (i.e. adverbials were not eliminated from the 
research). However, we made a decision to examine some other factors that affect the frequency 
of this transfer rather than adverbials. $e reason for that choice lies in the fact that we observed 
(see Section 1) that the majority of linguists puts a great deal of emphasis on adverbials. As far as 
adverbials are concerned, our results are similar to those of the above mentioned linguists. $e 
major aim of this article is to examine this transfer from a slightly different perspective. Hence 
it follows that this article also deals with other factors which are rarely mentioned in linguistic 
literature. Our article thus provides a new insight into this phenomenon by treating it from a 
slightly different perspective. 
$is section is divided into seven sub-sections. Hence it follows that we examine seven factors 
that may affect the frequency of this transfer.
 BrE AmE
no. % no. %
present perf. 314 87.2 232 64.4
preterite 46 12.8 128 35.6
Sum 360 100 360 100
perf.:pret. 6.8 1.8
χ
2
   s.l.> 99 %


0,0
20,0
40,0
60,0
80,0
100,0
BrE AmE
varieties
percentage
present perfect
preterite
 $e data in Table 1 show that the colloquial preterite takes over the functions of the present 
perfect in both varieties. $e difference occurs in the frequency of this phenomenon. We try 
to find whether the factor of the variety affects the frequency of the transfer of the functions 
of the present perfect into the domain of the colloquial preterite by using the chi-square test 
(henceforth: χ
2
−test). $e variety proves to be a statistically significant factor (with more 
than 99% significance level) that affects the frequency of the transfer, i.e. the transfer is more 
frequent in AmE than in BrE. $ese findings confirm the findings of the other research into this 
phenomenon as those of e.g. Dušková (1976), Sheen (1994) and Elsness (1997). $e colloquial 
preterite which behaves as a linguistic variant of the present perfect is illustrated in the following 
examples taken from our corpora:
(1) BrE
You mean it’s you buying the house? It’s all you, Marion, is it? I always hated you. You horrible 
bitch, you cunt cunt cunt-
(Churchill 1973)
(2) BrE
I was never a lazy girl, Marion tries hard. I work like a dog. Most women are fleas but I’m the 
dog.
(Churchill 1973)
(3) BrE
So you read the treatment?
(Elton 1991)
(4) AmE
We just got here.
(Shepard 1978)

(5) AmE 
My television is dead…My television…Is this Sandy?…My television died.
(Simon 1972)
(6) AmE 
Listen, I changed my mind. I’m not doing it.
(Simon 1972)
It is worth mentioning that the above constructions are entirely replaceable by the present perfect 
from the temporal-referential point of view, e.g. the construction I always hated you is replaceable 
by I’ve always hated you, i.e. the temporal meaning is the same irrespective of the use of the tense 
or the construction So you read the treatment? by So you’ve read the treatment?, My television died 
by My television has died and Listen, I changed my mind by Listen, I’ve changed my mind, etc.
$e colloquial preterite constructions fulfil all the criteria of the temporal-referential structure 
of the present perfect. However, we do emphasize that these constructions have unchanged 
meaning only from the strict temporal-referential perspective. $e different connotations that the 
colloquial preterite denotes belong to an entirely different domain that has nothing in common 
with the temporal location.
 At this point it is necessary to provide the definition of the indefinite
2
 and continuative 
interpretation of the present perfect. $e indefinite present perfect locates the situation in the 
period that reaches up to the time of utterance (henceforth: TU), whereas the situation itself lies 
wholly before TU according to Declerck (1991, 28). $is interpretation does not require any 
explicit reference to a period that reaches up to TU. In indefinite interpretation this reference 
is usually made by the present perfect itself, which indicates that the period leads up to TU 
(e.g. John has lived in Paris → indefinite interpretation → no explicit reference). We consider 
indefinite interpretation to be more salient because it requires no explicit reference, whereas 
the continuative usually requires an explicit reference to a period that reaches up to TU. $e 
continuative perfect indicates that the situation that is located in the period that reaches up to 
TU itself leads up to TU (e.g. John has lived in Paris for some time → continuative interpretation 
→ explicit reference established by for some time).
We check the claim that the transfer especially occurs in the indefinite meaning, i.e. we check 
whether the meaning is a statistically significant factor that affects the frequency of the transfer 
in both varieties by using the χ
2
−test. $e results are as follows:
 The majority of the above cited linguists do not use the concept of indefinite meaning, but they make mention of the resultative and 
experiential perfect, which roughly corresponds to the concept of the indefinite meaning.

British corpus
indef. cont. other
3
sum
no. % no. % no. % no. %
Br. perf. 236 85.8 71 91 7 100 314 87.2
Br. pret. 39 14.2 7 9 0 0 46 12.8
sum 275 100 78 100 7 100 360 100
perf.: pret. 6.1 10.1 - 6.8
χ
2
n.s.
 χ
 American corpus
indef. cont. other sum
no. % no. % no. % no. %
am. perf. 129 55.1 100 81.3 3 100 232 64.4
am. pret. 105 44.9 23 18.7 0 0 128 35.6
sum 234 100 123 100 3 100 360 100
perf.: pret. 1.2 4.3 - 1.8
χ
2
s.l.> 99%
 χ
 $e test reveals that the factor of meaning is not of statistical significance in BrE. Despite this 
fact, the transfer is more frequent in the indefinite meaning. Conversely, it is less frequent in the 
continuative meaning. In AmE the factor of meaning is significant. It is more frequent in the 
indefinite meaning, whereas it is less frequent in the continuative meaning. 
$e above given data confirm the claims of some linguists that this transfer is especially frequent 
in the indefinite meaning. Meaning is a statistically significant factor that affects the frequency of 
the transfer in AmE, whereas this factor is insignificant in BrE – though this variety also shows 
the tendency of more frequent occurrence of the colloquial preterite in the indefinite meaning. 
$is finding is logical if it is enlightened from the perspective of the different temporal-referential 
structure of the present perfect and traditional preterite.


We use the χ
2
−test to check whether the factor of regular and irregular verbs significantly affects 
the frequency of the transfer in both varieties:
British corpus
regular verb irregular verb indeter. sum
no. % no. % no. % no. %
Br. perf. 86 78.9 226 90.8 2 100.0 314 87.2
Br. pret. 23 21.1 23 9.2 0 0.0 46 12.8
sum 109 100.0 249 100.0 2 100.0 360 100.0
perf.: pret. 3.7 9.8 - 6.8
χ
2
s.l.> 99%
 χ
 $e test reveals that in BrE the factor of regular and irregular verbs significantly affects the 
transfer, i.e. the transfer is more frequent in regular verbs than in the whole corpus, whereas it is 
less frequent in irregular verbs.
American corpus
regular verb irregular verb sum
no. % no. % no. %
Am. perf. 71 61.2 161 66.0 232 64.4
Am. pret. 45 38.8 83 34.0 128 35.6
sum 116 100 244 100 360 100
perf.: pret. 1.6 1.9 1.8
χ
2
n.s.
            
 χ
 $e test shows that in AmE the factor of regular and irregular verbs has no statistically significant 
role in the transfer although the same tendency can be noticed as in BrE. $e transfer is more 
frequent in regular verbs than in the whole corpus, whereas it is of less frequent occurrence in 
irregular verbs.

$e results are surprising because the factor of regular and irregular verbs is the only one that affects 
the transfer in BrE only, whereas in AmE it is insignificant. $e percentage indicates that there 
is the same tendency in both varieties. $e transfer is more frequent in regular verbs. $e reason 
may lie in the fact that all regular verbs have the identical form of the preterite and past participle. 
Nevertheless, we believe that the checking the effect of the factor of the identical and different 
forms is far superior because the preterite and past participle of some irregular verbs are identical in 
form, whereas the others are different. We examine this factor in the following Sub-section.
 In this sub-section we test Elsness’s finding (1997, 235–6) that the transfer is more frequent 
when the forms of the preterite and past participle are identical, whereas it is less frequent when 
the forms are different. According to him, this is typical of both varieties, but the tendency is 
stronger in AmE. We examine whether the factor of the identical and different forms significantly 
affects the transfer in both varieties by using the χ
2
−test. $e results of our test confirm Elsness's 
findings. $e test reveals that the identical and different forms of the preterite and past participle 
is a statistically significant factor (with more than 95% significance level) in BrE, which means 
that the transfer is more frequent when the forms are identical, whereas it is less frequent when 
the forms are different.
British corpus
identical different other sum
no. % no. % no. % no. %
Br. perf 177 84.3 135 91.2 2 100 314 87.2
Br. pret. 33 15.7 13 8.8 0 0 46 12.8
sum 210 100 148 100 2 100 360 100
perf.: pret 5.4 10.4 - 6.8
χ
2
s.l.> 95%
 χ
 − In AmE the results are similar as in BrE. Identical and different forms of the preterite and 
past participle is a statistically significant factor that affects the transfer with more than 99% 
significance level. $e transfer is more frequent when the forms are identical. $is factor has a 
statistically significant role in both varieties, but the difference appears in the level of reliability. 
In AmE this factor is significant, with more than 99% significance level, whereas in BrE it is so 
with more than 95% significance level. $is finding is in accordance with Elsness’s, who claims 
that this tendency is stronger in AmE.

American corpus
identical different sum
no. % no. % no. %
Am. perf 123 54.7 109 80.7 232 64.4
Am. pret. 102 45.3 26 19.3 128 35.6
sum 225 100 135 100 360 100
perf.: pret. 1.2 4.2 1.8
χ
2
s.l.> 99%
               χ
 − Our results thus confirm Elsness’s findings and reveals that this factor is of statistical significance 
not only in AmE but also in BrE. In our view, these results are the proof that the distinction 
between the present perfect and traditional preterite is problematic even among native speakers. 
$e preterite is chosen only because of the identical forms of the preterite and past participle, i.e. 
for practical reasons, which are formally grounded.
 As Elsness (1997, 236) claims, the transfer occurs more frequently in the affirmative form, 
whereas it is less frequent in both the negative and interrogative form. He believes that the 
transfer occurs less frequently in the negative and interrogative form because the auxiliary appears 
in its uncontracted form also in informal language. On the other hand, the transfer occurs more 
frequently in the affirmative form. In this form the uncontracted auxiliary appears infrequently 
in informal language. $e uncontracted auxiliary have/has in the affirmative form would be very 
unusual in spoken language, whereas in the contracted auxiliary it is almost inaudible, this being 
the reason for the preference of the colloquial preterite in both varieties.
British corpus
affirm. negative inter. other sum
no. % no. % no. % no. % no. %
Br. perf. 232 86.2 37 94.9 43 86 2 100 314 87.2
Br. pret. 37 13.8 2 5.1 7 14 0 0 46 12.8
sum 269 100 39 100 50 100 2 100 360 100
perf.: pret. 6.3 18.5 6.1 - 6.8
χ
2
n.s.
 χ
 −

$e test reveals that the factor of form of the verbal phrase is insignificant in BrE, but the transfer 
is more frequent in the affirmative and interrogative form, whereas it is less frequent in the 
negative form.
American corpus
affirm. negative inter. other sum
no. % no. % no. % no. % no. %
Am. perf. 163 62.0 41 87.2 27 55.1 1 100 232 64.4
Am. pret. 100 38.0 6 12.8 22 44.9 0 0 128 35.6
sum 263 100 47 100 49 100 1 100 360 100
perf.: pret. 1.6 6.8 1.2 - 1.8
χ
2
s.l.> 99%
 χ
 − $e results are different in AmE, in which the form of the verbal phrase is a statistically significant 
factor. $e transfer is more frequent in the affirmative and interrogative form than in the whole 
corpus, whereas it is less frequent in the negative form. $e difference between BrE and AmE 
occurs in the fact that this factor is of no statistical significance in BrE, whereas in AmE it is 
significant. Nevertheless, the tendency is the same in both varieties.
Our results are not in accordance with those of Elsness. Our test does confirm that the transfer 
is more frequent in the affirmative form in AmE, but it also reveals that it is even more frequent 
in the interrogative form. $us the results show that the transfer occurs more frequently in 
the affirmative and interrogative form. $e transfer is less frequent in the negative form. 
Unfortunately, we are not able to answer why the transfer also occurs in the interrogative form. 
 Number 
Elsness’s findings (1997, 197–8) show that the ratio between the present perfect and traditional 
preterite is higher with plural than with singular subjects. We test whether number has a 
statistically significant role in the transfer. $e test reveals that number is not a significant factor 
in either BrE or AmE. $ere is the same tendency in both varieties. $e colloquial preterite 
occurs more frequently with singular subjects both in BrE and AmE, whereas it occurs less 
frequently with plural subjects. 

British corpus
sg. unct. pl. sum
no. % no. % no. % no. %
Br. perf. 253 86.3 7 87.5 54 91.5 314 87.2
Br. pret. 40 13.7 1 12.5 5 8.5 46 12.8
sum 293 100 8 100 59 100 360 100
perf.: pret. 6.3 7.0 10.8 6.8
χ
2
n.s.
 χ
 − American corpus
sg. unct. pl. sum
no. % no. % no. % no. %
Am. perf. 213 64 2 100 17 68 232 64.4
Am. pret. 120 36 0 0 8 32 128 35.6
sum 333 100 2 100 25 100 360 100
perf.: pret. 1.8 - 2.1 1.8
χ
2
n.s.
 χ
 − Number is insignificant in each variety, but the tendency towards the more frequent transfer in 
singular subjects occurs in both varieties. $ese results can be explained by the Naturalness $eory 
as formulated by Orešnik (1999). According to his claim, the present perfect is a compound 
tense and as such it represents a strong variant. It is typical of a strong variant that it first prevails 
in less natural environments. On the other hand, the preterite is not a compound tense, hence a 
weak variant, which first prevails in more natural environments. As far as number is concerned, 
Orešnik (1999, 91) claims that the singular is ‘less marked’ or ‘more natural’ than non-singular. 
Hence it follows:
>naturalness (singular, non-singular); after Mayerthaler 1987, 49.
$is theory enables us to explain why the colloquial preterite occurs more frequently with singular 
subjects, i.e. since the preterite is not a compound tense, it prevails in more natural environments 
(i.e. singular). $e present perfect, on the other hand, is a compound tense. As a consequence it 
prevails in less natural environments (i.e. non-singular).

Person
Elsness (1997, 197−201) finds that the ratio between the present perfect and traditional preterite 
is highest for the first and the second person, whereas it is lowest for the third person. According 
to his opinion (1997, 198), the reason lies in the fact that the first and the second person usually 
express extra-textual, situational reference, whereas the third person expresses anaphoric textual 
reference. Hence it follows that the traditional preterite is more frequently used with the third 
person. Our hypothesis is that the colloquial preterite behaves as if it were the present perfect and 
is therefore more frequently used with the first and the second person. $e test reveals that person 
is insignificant in either variety. But varieties have different tendencies. In BrE the transfer is less 
frequent with the first and the second person, whereas it is more frequent with the third person.
British corpus
1
st
 pers. 2
nd
 pers. 3
rd
 pers. sum
no. % no. % no. % no. %
Br.perf. 146 88 66 89.2 102 85 314 87.2
Br. pret. 20 12 8 10.8 18 15 46 12.8
sum 166 100 74 100 120 100 360 100
perf.: pret 7.3 8.3 5.7 6.8
χ
2
n.s.
 χ
 − $e situation in AmE is different. Although person is not a significant factor, the transfer is the 
most frequent with the first person, whereas it is less frequent with the second and the third 
person.
American corpus
1
st
 pers. 2
nd
 pers. 3
rd
 pers. sum
no. % no. % no. % no. %
Am.perf. 102 62.2 62 66 68 66.7 232 64.4
Am. pret. 62 37.8 32 34 34 33.3 128 35.6
sum 164 100 94 100 102 100 360 100
perf.: pret. 1.6 1.9 2 1.8
χ
2
n.s.
 χ
 −

$e results show that the transfer occurs most frequently with the third person in BrE, 
whereas in AmE it is with the first person, which confirms our expectation. We believe that 
these differences are not necessarily a reflection of the actual differences between the varieties 
because this distribution may be due to extra-linguistic factors such as the type and nature of 
conversations, which may dictate the use of person and number. It is true that the third person 
expresses anaphoric reference more frequently, but this is not necessarily always true.
 We check the effect of the type of clause on this transfer. We want to find whether the type of 
clause is a factor that has a statistically significant role in the frequency of the transfer by using 
the χ
2
−test. $e results are as follows:
British corpus
independent depend. indeter. sum
no. % no. % no. % no. %
Br. perf. 227 85.7 85 91.4 2 100 314 87.2
Br. pret. 38 14.3 8 8.6 0 0 46 12.8
sum 265 100 93 100 2 100 360 100
perf.: pret. 6.0 10.6 6.8
χ
2
n.s.
 χ
 − $e test reveals that the type of clause is not a statistically significant factor in BrE, but the results 
indicate that the transfer is more frequent in independent clauses, whereas it is less frequent in 
dependent clauses.
American corpus
independent depend. sum
% no. % no. %
Am. perf. 176 61.5 56 75.7 232 64.4
Am. pret. 110 38.5 18 24.3 128 35.6
Sum 286 100 74 100 360 100
perf.: pret. 1.6 3.1 1.8
χ
2
s.l.> 95%
 χ
 −

$e test reveals that the type of clause is a statistically significant factor in AmE. $e transfer is more 
frequent in independent clauses, whereas it is of less frequent occurrence in dependent clauses.
$e factor of the type of clause has a statistically significant role in the frequency of the transfer in 
AmE. $is factor is of no statistical significance in BrE although the tendency is the same as in AmE. 
In both varieties the transfer is more frequent in independent clauses than in dependent ones.
Our results can be made explicable by means of the Naturalness $eory (Orešnik 1999). As 
Orešnik states, the present perfect is a strong variant and prevails in less natural environments. 
On the other hand, the preterite is a weak variant, which prevails in more natural environments. 
As far as a clause is concerned, Orešnik (1999, 23) claims that an independent clause is ‘more 
natural’ than a dependent clause. Hence it follows:
>naturalness (independent clause, dependent clause). 
It is therefore obvious why the transfer is more frequent in independent clauses. $e preterite 
is a weak variant and tends to occur in more natural environments and independent clauses are 
more natural than dependent ones.
 $e main finding of our research was that the factor of the variety has a statistically significant 
role in the transfer of the functions of the present perfect into the domain of the preterite, i.e. 
this transfer is more frequent in AmE than in BrE.
Factor Variety Tendency
BrE AmE
meaning: indefinite  
and continuative
n.s. s.l. > 99% same
verbs: regular and irregular verbs s.l. > 99% n.s. same
identical and different forms of 
the preterite and past participle
s.l. > 95% s.l. > 99% same
form of the verbal phrase: 
affirmative, negative and 
interrogative
n.s. s.l. > 99% same
number of the person: singular  
and plural
n.s. n.s. same
person: 1
st
 person, 2
nd
 person  
and 3
rd
 person
n.s. n.s. different
type of the clause: independent  
and dependent
n.s. s.l. > 95% same


$e results of our research showed that the significant similarity between BrE and AmE (see T able 
16) in the transfer occurs only in one factor, i.e. in the factor of identical and different forms of the 
preterite and past participle. $is factor significantly affects the transfer in both varieties, i.e. the 
transfer is more frequent when the forms are identical and infrequent when the forms are different. 
$e research showed that the differences between BrE and AmE in this transfer mostly occur: (1) 
in the frequency of this transfer, (2) in the number of the statistically significant factors that affect 
the frequency of the transfer, (3) in the factor of meaning (indefinite and continuative), (4) in the 
factor of form of the verbal phrase (affirmative, negative and interrogative), (5) in the factor of 
verbs (regular and irregular), and (6) in the factor of type of clauses (independent and dependent 
clauses). $ere are only two significant factors that affect the frequency of this transfer in BrE, 
whereas there are more significant factors in AmE. $e factor of meaning is not significant in 
BrE, whereas it is of statistical significance in AmE. $is transfer is more frequent in indefinite 
meaning and infrequent in continuative meaning. Notwithstanding that the factor of meaning 
is not significant in BrE, the transfer is more frequent in indefinite meaning. $e factor of 
form of a verbal phrase does not significantly affect the transfer in BrE. In AmE this factor is of 
statistical significance, i.e. the transfer is more frequent in affirmative and interrogative form, 
whereas in negative form it is infrequent. $ere is the same tendency in BrE. $e factor of regular 
and irregular verbs significantly affects the transfer in BrE, i.e. the transfer is more frequent in 
constructions with regular verbs. In AmE this factor is not significant, but the tendency is the 
same. $is factor is the only one that is significant in BrE but is of no significance in AmE. $e 
factor of type of clauses (dependent and independent clauses) is not significant in BrE, whereas 
it is of significance in AmE, i.e. the transfer is more frequent in independent and infrequent in 
dependent clauses. $ere is the same tendency in BrE.
$e factors that have no significant role in the frequency of this transfer either in BrE or in AmE 
are (1) the factor of number of the person (sg. and pl.), and (2) the factor of person (1
st
, 2
nd 
and 
3
rd
 person).
In conclusion, although there is a considerable difference between BrE and AmE in the transfer 
of the present perfect functions into the domain of the colloquial preterite, we believe that both 
varieties are moving in the same direction and that the British variety is under increasing influence 
of AmE. Nevertheless, the general tendency is the same in both varieties. $is fact indicates 
that this phenomenon, which was first noticed in informal American English, is spreading to 
informal British English. $e question of whether or when this phenomenon enters the formal 
varieties cannot be answered for the time being.
 
 Comrie, B. 1976. . Cambridge: Cambridge University Press.
Crystal, D. 1995. . Cambridge: Cambridge University Press.
Declerck, R. 1991.  London: Routledge.

Dušková, L. 1976. On some differences in the use of the perfect and the preterite between British and American English.  5: 53–68.
Elsness, J. 1997 . Berlin and New York: Mouton de Gruyter.
Ferligoj, A. 1995. . Ljubljana: Tiskarna Mišmaš.
Greenbaum, S. 1996. Oxford: Oxford University Press.
Leech, G.N. 1987 . . London and New York: Longman.
Leech, G., and J. Svartvik. 1994. London and New York: Longman.
Marshall, H.W. 1979. The colloquial preterite versus the present perfect. PhD diss., Teachers’ College, Columbia University: 
University Microfilms International.
McArthur, T., ed. 1992. Oxford and New York: Oxford University Press.
Orešnik, J. 1999. . Ljubljana: Slovenska akademija znanosti in umetnosti.
Peterson, B.A. 1970. Towards Understanding The “Perfect” Constructions in Spoken English.  8 (1): 2–10.
Quirk R., S. Greenbaum, G. Leech, and J. Svartvik. 1985. London and New 
York: Longman.
Richards, J.C. 1979. Introducing the Perfect: An Exercise in Pedagogic Grammar. 13 (4): 495–500.
Sheen, R. 1984. Current Usage of the Simple Past and Present Perfect and Its Relevance for Teaching English as a Foreign or 
Second Language.  40 (3): 37 4–85.
Sheen, R. 1994. A Neglected Difference between British and American English. 1: 45–7 .
Vanneck, G. 1958.  The Colloquial Preterite in Modern American English. 14: 237–24.
Žetko, K.D. 2003. The present perfect in modern British and American English. Master’s thesis, University of Ljubljana.