The Making of the English Verb-System


    Traditional comparative grammar did not see different ethnic layers and explained the development from Indo-European to Modern English as a series of inner sound shifts within one language. A new look at Common Germanic demonstrated that Grimm’s sound shifts (Lautverschiebungen) had parallels in Armenian and could be due to ethnic mixing (Gamkrelidze, Ivanov 1982). Their natural account may explain them as an exchange of loanwords between Viking fishermen and Anglo-Saxon peasants. Table 28 demonstrates these contacts as ‘two-way projections’ making mutual imprints on two overlapping dialects. The chief problem consisted in the Viking (Palaeo-Turcoid) opposition of fortes and lenes that underwent aspiration in initial position and gemination in medial positions. According to this law, fricatives remained voiceless in initial positions but exhibited voicing and gemination in medial positions (thin /θ-/ vs. leather /-ðð-/). The Viking initial fortis sonants were imprinted into Old English as sounds with a strong pre-aspirated explosion (OE hlāf ‘loaf’, hrōf ‘roof’ and hnutu ‘nut’). On the other hand, words of Indo-European (Anglo-Saxon) origin preserved non-aspirated initial sonants (OE mōdor ‘mother’, niht ‘night’). The Viking word-stock can be seen in all words with pre-aspirated consonants hp, ht, hk, hm, hn, hl, hw while the Indo-European word-stock was remarkable for pre-assibilated clusters sp-, st-, sk- sm-, sn-, sl-, sw-. These initial clusters were taken over from Indo-European dialects but they were due to earlier receptions of initial fortes from Mesolithic hunters. They probably arose from assibilating foreign pre-aspirated consonants hp, ht, hk, hm, hn, hl, hw.  There were no one-way shifts within one system but only mutual imprints of overlapping languages producing ‘two-way translations’ into heterogeneous phonologies.

   Classical historical grammar believed in lawful sound shifts operating as an imaginary clock on one national literary standard. It remained blind to numerous spoken tribal subcomponents that dominated or succumbed according as their speakers and kinsmen succeeded in social and military competition. When the Wessex king Egbert conquered Mercia in 829, England united also in using the Wessex literary standard, but its assumed sound shifts only changed the mutual hierarchy of spoken regional dialects. Languages do not evolve from their own will and needs but in accordance with the social and geographic possession of their speakers. The historical diagram on Table 29 demonstrates the linguistic evolution of English as a variable dependent on ethnic migrations and conquests. It was not a story of one united nation but of incessant mutual clashes between tribes of different origin. Besides the Fist Northern Culture of arctic fishermen there was the Indo-European Battle-Axe People and Scots as heirs of the Megalith culture (3,200 BC) coming from Spain. Anglo-Saxons conquered Celtic Britain and subdued the autochthonous populations of Britons and Gaels (Gaels – Goidels - Gwynt – Albans) but they ‘Englished’ Britain only at cost of being ‘Britonised’ by the absorbed Britons. The Norman Conquest resulted in a partial ‘Normanisation’ of Middle English though its effects were clearly seen only after centuries when English gradually ‘re-Englished’ and the Norman impact weakened. New English emerged under the Tudors when London merchants seized the rule and expropriated the literary standard from Lancaster and Yorkshire landowners. The Celtic brachycephalic Gaels kept silent for centuries as artisans and small townsmen but they raised their heads during the Puritan Revolution in 1640. They seized the Parliament as the Puritan Roundheads and promoted their popular speech with many Celtic survivals to the official standard. What looked like sound shifts and consonant laws was actually inner reshuffling between social layers and military castes.

     The basic stages in the evolution of English are seen on Table 30 displaying how its tenses and moods composed from three different ethnic components. One subgrammar was due to the old Anglo-Saxons whose system dominated in Old English and was partially restored again in Early New English.

    The second stage started with the Norman Conquest and Anglo-Norman French whose reign melted Middle English into an analytic language with verb phrases combining auxiliaries with non-finite verb-forms. The new system was based on analytic perfects composed from the auxiliary to be and a past participle. Such constructions are evidenced in ME lenten is cumen ‘spring has come’, Estonian olen lugenud ‘I have read’ and Turkish sevdí idim ‘I have loved‘. The analytic layout of the ME verbal system was strengthened also by the analytic future tense and conditionals. These began to compete with OE subjunctives and replace them in all positions except for conditional clauses.

The analytic verbal systems enforced constructions of auxiliary verbs with non-finite verb forms, participles, gerunds and infinitives. Middle English took the gerundial construction over from the Anglo-Norman gérondif and adapted infinitives from Old English verbal nouns. Both forms are typical of Turcoid and Ural-Altaic languages where they function as a makeshift for hypotactic subordination and that-clauses. In Indo-European, Caucasoid and Bantu languages there is a strong tendency to use hypotactic that-clauses and apply subjunctives as special tenses for that-clauses. Under the Norman influence Middle English became reluctant to that-clauses and began to replace the clause She commands that he be obedient by the accusative-with-infinitive construction She commands him to be obedient. I. Poldauf (1958: 177) described this tendency as ‘secondary predication’, J. Hladký (1961: 105ff.) as ‘condensation’ and L. Dušková (1988: 542) as ‘semipredication‘.

  The Norman rule confined the Anglo-Saxon and Celtic component to popular speech but new democratic changes made it emerge again in New English. Common townsmen infiltrated English grammar with remains of Celtic morphology, especially with progressive tenses (I am hunting) and ‘predicatives of state’ (I am afraid, We are aboard). They were taken from Celtic languages and through transitional forms I am a-hunting, I am on fright in popular speech they paved their way into the literary standard. The Puritan Round-heads began to use them in literary English and build its grammar on the opposition of simple and progressive tenses. The simple present adopted the auxiliary do and began to function as the Celtic habitualis (I do not write). This auxiliary stemmed from the Old English modal verb dugan ‘to avail, zu taugen’ and had the pronunciation he doth /daq/ know different from the full-meaning verb he doeth /du:iq/ nothing. Its counterpart was the actualis (I am writing) denoting presently proceeding actions. The progressive present ‘I am allowing’ reads in Modern Irish táig ag ligean, in Gaelic tha mi a' leigeil and in Manx ta mee lhiggal (Lockwood 1975: 107ff.). Outside the British Isles the habitual and progressive tenses can be seen only in Albanian, which has two progressives, Po(punoj) and Yam tue punue ‘I am writing’ (Ejntrej 1982: 84). The same choice of Palaeo-Gallic languages applies parallels to the English English. Its present and past tense stood in opposition to subjunctive optatives applied after conjunctions and in that-clauses. There were no future tenses, no conditionals, almost no perfects, no gerunds and no progressive tenses. Now English exhibits a monstrous system of almost fifty tenses and moods but its huge structural complexity has grown from very simple elementary origins. ‘immediative future tense’ (It’s going to rain) and phrases denoted as ‘predicatives of state’ (She stood aghast).


Table 1 The linguistic prehistory of British and Germanic nations







he is

he be

Flowchart: Delay:  Norman
that he be


he was

he were

that he were





middle class


present imperfect

he is

he will be                     

Flowchart: Delay:   Celtic
present perfect

he has been      

he will have been

past (imperfect)

he was

he would be                           

past perfect

he had been

he would have been





present (imperfect)


being asked

past (perfect) gerund

having asked

having been asked






I do not ask

I am not asking  


I  will ask

I am going to ask


I  used to ask

I was asking


Table 2 Different ethnic layers in the English tense system

    Modern English represents a live amalgam of at least three subgrammars with several vital pure tenses but also many hybrids or changelings. Hybrid subjunctives such as She have come or He were reading are doomed to die because they mix forms due to the Anglo-Saxon, the Norman and the Celtic subgrammar. On the other hand, the simple present tense They don’t play chess is bound to serve in several incoherent functions: as a Celtic habitualis, as a Norman imperfect present and an Anglo-Saxon praesens realis. English philology needs a tenable nomenclature acknowledging an inner hidden diversity of grammatical subsystems but discarding all secondary derived hybrids. It should be aware of the competition of several grammatical archetypes operating in modern Germanic languages, their structural coherence and typological diversity. It should admit that overlapping languages soak with isolated loanwords through neighbouring dialects and transplant into their soil also their phonetic and grammatical peculiarities. When adopting the Scottish place name Loch Ness, English tends to take over the Scottish phoneme X, and when borrowing the Latin loanword senior, English has to apply the syntax of   Latin comparatives (He is senior to her vs. He is older than her). At a definite level of quantitative growth such osmosis (soaking through) results in mutual imprints of subphonologies and subgrammars into the ruling literary standard.

    The functional core of Modern English still rests on the Norman subgrammar that may be reconstructed as the structural Urform of most Ural-Altaic languages. A more adequate taxonomy of its tenses operating in English might speak of the present imperfect (she goes), past imperfect (she went), present perfect (she has gone) and past perfect (she had gone). The opposition of perfects and imperfects operates also in the category of mood that suffers much from the misnomer ‘future tense’. F. A. Palmer, J. Lyons, G. N. Leech and R. Quirk refused to consider the English future tense as an indicative tense and proposed to regard it as a sort of mood. This form should be conceived as a form of unreal mood related closely to conditionals and called properly ‘future mood’, ‘predictive conditional’ or ‘real predictive’. Their correspondence becomes apparent when If I come I will see in real (open) conditions is shifted into If I came I would see in unreal (hypothetical) conditions. However, it is not convenient to join some authors in calling would do a ‘preterit’ from will do, we had better call the former ‘unreal predictive’ and the latter ‘real predictive’ because they convey prediction. Then we would be free to re-classify the English mood forms as the real imperfect predictive (she will go), real perfect predictive (she will have gone), unreal imperfect predictive (she would go) and unreal perfect predictive (she would have gone).

     Further inconsistencies are found in non-finite verb-forms exhibiting no symmetry to finite verb-forms. Semi-predicative verb-forms deserve taxonomy compatible with finite tenses because the correlation between gerunds and infinitives corresponds to that between indicative and conditional (predictive) mood. This incoherent usage might be corrected by introducing the pair of ‘finitivals’ and ‘infinitivals’. Finitivals would cover all finite tenses while infinitivals would include all non-finite verb-forms. Their tenable taxonomy in English might consist of the imperfect indicative infinitival (our doing), perfect indicative infinitival (our having done), imperfect conditional infinitival (to do) and perfect conditional infinitival (to have done).




present imperfect

present indicative

future predictive


he asks

he will ask                     

present perfect

pre-present indicative

pre-future predictive


he has asked

he will have asked 

past (imperfect)

pre-preterit indicative

pre-conditional predictive


he asked

he would ask                           

past perfect

pre-preterit indicative

pre-conditional predictive


he had asked

he would have asked 




present (imperfect)  

present ind. infinitival

our asking

present pred. infinitival

to ask


pre-present ind. infin.

pre-present pred. infin.


our having asked

to have asked

Table 3 A systematic taxonomy of English verb-forms

     Such terms might get a chance in academic grammars but they are unlikely to domesticate in live school usage. Live usage will always tend to omit loci communes and drop futile attributes such as ‘indicative’ or ‘imperfect’ A compromising solution might replace the redundant perfect/imperfect correlation by the pair of ‘preterit’ and ‘pre-preterit’. Such reformed or rationalised nomenclature of English verb-forms is suggested in Table 31. In its proposal We will have written would be referred to as ‘pre-future predictive’ and She would have brought as ‘pre-conditional predictive’. Redundant terms may be deleted by preserving distinctive attributes in ‘marked categories‘ and dropping them in ‘non-marked categories’ (Prague School coinage). Then the cumbersome term ‘pre-present predictive infinitival’ denoting to have asked could be reduced to ‘pre-present infinitival’ and the ‘pre-present indicative infinitival’ denoting our having asked would shrink to ‘present par-infinitival’. Such changes would meet requirements of both structural symmetry and easy practical reference.


Extract from Pavel Bělíček: The Making of the English Tense System. In: The 6th Conference of British, American, and Canadian Studies. Opava: Slez. Univ.  2001, pp. 2-8.