English is a Germanic language. So too is German and Dutch. While geographically separated by the English Channel, these languages generally have rather similar histories, up until several centuries ago.

If you have learned these languages before, you would find that many German and Dutch participles contain the prefix ge-, while English participles lack this feature. Was this always the case, and if not, when and how did English lose it?

The short answer is, no. Old English generally had a similar structure of participles with German and Dutch even until the Middle Ages, although that would have changed quite a bit in Middle English. It was in Modern English when this prefix was fully lost, except in particular exceptions.

One notable exception is the verb “to clepe”, which means to give someone or something a specified name. While an archaic expression, its participle interestingly includes “yclept”, a relic of the participle system in Middle English. An example of its usage can be found in The Monctons: A Novel written by Susanna Moodie:

A certain impetuous, wilful, wrong-headed boy, yclept Geoffrey Moncton.

The Monctons: A Novel, by Susanna Moodie

Perhaps a more commonly used example is the word “enough”, which derives from the Old English word genög, which had that ge- prefix, but is now reduced to e-. Its Modern German counterpart? Well, that will be genug, which preserves this ge- prefix entirely.

There are several key problems historical linguistics encounter when trying to pinpoint the exact period in which this ge- prefix was lost in the English language. For one, we cannot precisely know when a surviving piece of text was written. And that is assuming we have encountered enough surviving manuscripts to begin with. Some corpus data split these surviving texts up into year bands of up to a century long, and so one piece of text might be composed at some time in the years 850 – 950, while another might be written in 1050 – 1150. The most precise date range one could give is in the span of several decades.

The second problem is the problem of confounding introduced by the Norman conquest of England in 1066. To illustrate this problem, we will take a look at the York-Toronto-Helsinki Parsed Corpus of Old English Prose (YCOE). This corpus of Old English texts classifies Old English into four periods, namely, up to 850, from 850 to 950, from 950 to 1050, and from 1050 to 1150. When looking at the prevalence of the use of the ge- prefix in 1050 to 1150, we see a strange drop in the amount of corpus data points compared to previous year bands, as illustrated in this 2015 paper by McFadden:

Period	ge-	no ge-	% ge-
pre-950	15330	44721	25.5
950 – 1050	19865	58792	25.3
1050 – 1150	74	430	14.7

While on a crude analysis, one might be quick to see the prevalence of the prefix ge- drop towards the end of Old English, one should also consider why there are magnitudes fewer data points to work with for the period 1050 – 1150. This period coincided with the period of Norman conquest in England. As McFadden explained, prior to this conquest, there was a relatively consistent written form of Old English, but was quite conservative. This meant that over time, what was written in manuscripts was increasingly detached from what was actually said and used by speakers.

The Norman conquests of England basically did away with this consistency. From then on, people could write in a manner that reflected their own speech, rather than an orthographical standard. In a way, this did not imply that the systems of written or spoken English were heavily destroyed, but more rather, English writing was finally allowed to catch up several centuries of changes in English speech, grammar, and vocabulary.

So to circle back, what does this data tell us?

It seems to suggest that the usage of the ge- prefix in written Old English was fairly stable, although its decline started to be noticeable in the period coinciding with the conquest and aftermath of the Normans. Perhaps this decline in the use of the ge- prefix occurred over a more gradual period, and it took the Norman conquest to upend this scribal tradition, bringing Old English writing to reflect the decades and centuries of changes in the Old English language.

This is when we bring in more data points from other time periods, particularly the time of Middle English. This is the period in which the ge- prefix met its ultimate extinction, only to have some of it exist as “vestigial” structures in the words we use today. In McFadden’s study, he used data from the second edition of the Penn-Helsinki Parsed Corpus of Middle English (PPCME2), which works very similarly to the YCOE, but focuses on Middle English, defined as a period between 1150 and 1500. Time periods were divided into 1150 – 1250, 1250 – 1350, 1350 – 1420, and 1420 – 1500.

When data from this corpus are put together with those from YCOE, we start to notice a more prominent trend.

Period	ge-	no ge-	% ge-
pre-950	15079	43464	25.8
950 – 1050	19695	57793	25.4
1050 – 1150	74	422	14.9
1150 – 1250	2297	30190	7.1
1250 – 1350	989	16850	5.5
1350 – 1420	1106	58519	1.9
1420 – 1500	162	31614	0.5

At first glance, you might notice several differences between McFadden’s second table and the first. It could possibly be explained by newer data point entries made for Old English for the first table, or ambiguous or poor quality entries were removed for Old English in the second table. Accounting for this discrepancy was not really reported in this publication, and so this explanation could be just speculation at best.

But what we can essentially infer from this table, is that the usage of the ge- prefix has been in continual decline during the period Middle English was used. McFadden noted later on that the ge- prefix used in forming the passive voice and the perfect tense essentially slumped at some point in the 13th and 14th centuries, or the time corresponding to the second and third period of Middle English in the PPCME2, although the decline for the passive voice ge- was more steady than the perfect tense ge-.

Whether or not this decline actually started towards the end of Old English is not really clear, however. Additionally, this only documents the usage of the ge-prefix in spoken Old and Middle English, which might not adequately reflect the English spoken at those times. The ge- prefix might have started to decline earlier on, perhaps even in the 10th or 11th centuries, but orthography did not catch up with the changing spoken form. Putting a more exact date to the start of the decline of the ge- prefix is also next to impossible, due to the difficulties in attesting for the year a text was written.

Lastly, we want to try to answer how this ge- prefix was lost. As with the ge- prefix in languages like German, the ge- prefix in Old English is very likely unstressed. Instead of the /jɛ/ sound as one might expect it to be pronounced especially at the start of Old English, this would have been pronounced closer to /jə/ in Old English over time. Slowly but surely, this reduction would have progressed, giving us the /i. or /ə/ sounds over time. This is the reduction we see in the participle ‘yclept’, and in the word ‘enough’.

Some might point towards a more syntactical approach in understanding the decline of the ge- prefix. It turns out that the ge- prefix is not just used in forming participles in Old English; it also takes up other grammatical functions in Old English. However, the author noted that there must be some substantial change in Old English grammar that would show the redundancy of the ge- prefix in English over time.

Interestingly, a similar decline in the use of the ge- prefix was noted in the North Germanic languages, although the exact prefix might be a little different, like ga- or gi-. However, this loss very likely occurred at some point in prehistory, as Old Norse lacked this ge-, ga-, or gi- prefix. The ga- or gi- are proposed reconstructions for Proto-Germanic, the theoretical common ancestor of the Germanic languages.

A similar vein of reductions has been seen in some dialects of German and its cousins, or at least that was what I have seen in written Viennese German and the Bavarian language. Examples like gsagt (reduced from gesagt, ‘said’) come to mind, although we cannot really predict if the /g/ sound would entirely be lost over time. Perhaps with more research, we could build a comprehensive history of the prefix ge-, and how and why it has changed in the various languages that use it.

Further reading

McFadden, Thomas. 2015. “Preverbal Ge- in Old and Middle English”. ZAS Papers in Linguistics 58 (January):15-48. https://doi.org/10.21248/zaspil.58.2015.427.