The English language is notorious for its spelling, where there are a bunch of silent letters in many of its words, and how there is a many-to-many correspondence between many of its letters and sounds. From the ‘w’ in place names like Greenwich but not Sandwich to the inconsistencies in the pronunciation of the ‘-ough’ parts in ‘through’, ‘trough’, and ‘thorough’, to many who are learning English, young and old, such spelling inconsistencies have posed a challenge to many when it comes to reading and writing it. Today, we will talk about the silent letters in words like ‘knight’ and ‘sight’.

As many of us would know, the ‘k’ in words like ‘knee’, ‘know’, and ‘knight’ are not pronounced at all today. This is the same for the letter ‘g’ in words like ‘night’, ‘gnat’, and ‘furlough’. So, when did that become the norm, and why has spelling not managed to keep up with these sound changes?

In the centuries prior to the English we speak today, it turns out that these letters were indeed pronounced, as with some Germanic languages that are spoken today. One example is the name of the King of England from 1016 to 1035, King Cnut or Knútr. In Old Norse and Old and perhaps early Middle English, these ‘k’ and ‘n’ sounds would have been pronounced as a permitted consonant cluster.

Similarly, the letters ‘gh’ in ‘light’ and ‘sight’ would have been pronounced in Old and Middle English, but instead of the ending ‘-ight’ sound we are familiar with today, these would have been pronounced with a long ‘i’ vowel, /i:/, with the ‘gh’ being pronounced closer to a Dutch ‘g’ sound like /ɣ/ in Old English, and a /x/ or a /ç/ sound in Middle English. Compare these words with the German Licht and Sicht respectively.

In Middle English, some of these words that contain a ‘gh’ digraph today would have been written with an extinct letter, the yogh ȝ. This would be seen in words like wrouȝte (wrought) and niȝt (night), but also in words like ȝister– (yester-). With the use of the tailed ‘z’ though, the yogh would have faded away, and replaced with other letters like ‘gh’ and ‘y’, with Scots replacing some of them with a ‘z’ under confusion with the tailed counterpart (hence Menzies, from Menȝies).

These sounds were dropped at some point in the 16th century under a process of sound changes. This process is known as apheresis (AmE) or aphaeresis (BrE), where a word-initial sound is lost. Such a process has been documented in the development of English slang terms today. For example, the Cockney rhyming slang features some words that have undergone aphaeresis, such as ‘amsteads (Hampstead Heath) and ‘andsome (handsome). Thus, it could be the case that aphaeresis first arose as some sort of colloquial Middle English slang, and the unfavourability of pronouncing the ‘kn’ consonant cluster grew throughout the British Isles, leading to the omission of the ‘k’ sound we hear today. However, perhaps some French influence may have occurred as well, as French is not known to have such consonant clusters.

So, why has English spelling not caught up with these sound changes? English spelling as we know it for the most part started being set in the 14th century, with some sort of consistency or standardisation being set back then. Some letters were changed after the invention of the printing press and handwritten English (hence the fading of the yogh). It was right about when other major sound changes like the Great Vowel Shift were taking place, and so, many of these words we see today would have roughly indicated the pronunciation of such words back then, like night, from [ni:çt] to [naɪt], and sight, from [si:çt] to [saɪt]. Despite these sound changes, English spelling has mostly stuck with the standardisation that was adopted at that time. Over time, the English language grew to assimilate more words of foreign origin and international reach, perhaps making it increasingly difficult to exert large scale spelling reforms in ‘standardised English’.

However, in the centuries since, to say that there were no attempts at reform of English spelling would be grossly inaccurate. Some reforms led to the divergence between written American and British English, while others had some form of success, such as words of Greco-Latin origins like ‘island’, previously iland. This reform would have resulted the rise of the silent ‘g’ in words of Greco-Latin origins, like phlegm, from Greek phlegein (to burn), which changed from Middle English fleem, fleume. A similar case also applies for the word diaphragm, which has a silent ‘g’.

But for words like ‘knight’ and ‘gnat’, these spellings remained unchanged, perhaps presenting today as a vestige of what Old English and early Middle English would have sounded like, much like how the circumflex in French, in some cases, indicate the origins of some of its words like fenêtre (from fenestra).

Despite the general unfavourability of pronouncing the ‘kn’ consonant cluster in words like ‘knee’, the ‘kn’ cluster still persists when one talks about words of foreign origin, such as some surnames of Germanic origins like Knaus, or the Knesset (Israeli parliament). There are also some silent letters that I want to talk about, most notably the silent ‘b’ in words like ‘doubt’, ‘debt’, and ‘subtle’, so perhaps I will write another short essay about those in time to come.

Further Reading

O’g’li, S. B. T. (2021) ‘The role of shortening in the structural and semantic analysis of slang words’, Berlin Studies Transnational Journal of Science and Humanities, 1(16), pp. 18-29.