Welcome fellow colleagues! You surely remember my name - Jakob Grimm. I was briefly mentioned in the 2nd lecture where I was claimed to be one of the founders of comparative linguistics. Besides I'm most famous for the co-edition of the well-known Grimm's Fairy Tales, together with my dear brother Wilhelm. If you want to know more about the two of us, then have a look at this informative web site >here<.

Now, however, I am not going to debate one of those fascinating fairy tales. Our topic for today is the question of how the two words cannabis and hemp may be linguistically connected. Obviously both terms denote this common plant growing five-fingered leaves and which is quite famous for many reasons - facts that I will not intensify in this tutorial. For those of you wishing to have more information on the botanic, historic or narcotic background, I recommend this web site >here<.

Let us now return to the expressions in question. Lat. Cannabis is of Greek origin (kannabis). Hemp on the other side is a common Germanic word, as may be supported by Swed. hampa, Dutch hennep or by Germ. Hanf. Further attestations may be found in Persian kanab or in Sumerian kunibu.

So, by analyzing the phonemic consonant structure, we can set up three lists of corresponding records:

You do already guess what I am driving at! Right, my famous sound shift law, known as Grimm's Law, or First consonant Shift. For the sake of simplicity, I will redraw it here in it's very basic form:

And yes, you were right! All forms showing /k/ and /b/ (kannabis, kanab, kunibu) preserve the original PIE consonants. The second column with hemp, hampa and hennep exactly follows my law (k > h, b > p).

But what about Ger. Hanf? Well, if you contrast ModE. up with Germ. auf, or ModE. ship with Germ. Schiff you will find out some regularity in the way /p/ shifts to /f/ in German. This is explained by an other law, namely my Second Consonant Shift, which only affects High German and therefore is of lesser interest for the history of English.

So far, everything is comprehensible. Lat. cannabis and ModE. hemp are cognates, i.e. they descend both from a common PIE ancestor that we may reconstruct as something like *kannabiz. But looking at all these complex rules, you might ask yourself, what exactly was going on in far past of our language history. In more general terms, is there any explication, why phonetic clusters should shift at all and change their status?

In order to pursue this question, let us consider language as being a complex system. We are not concerned with any speakers, some isolated sounds, nor are we dealing with a specific language. The following considerations are kind of universal and solely rely on the basic structure of language, a system where everything is logically joined and where major changes do not occur in isolation but always in relation to each other. This approach is part of the structuralist paradigm and therefore the following explications are based on a structural change of the system.

To start, let us have a look at the (reconstructed) PIE sound system. For our purpose we will restrict the following considerations on obstruents, i.e. sounds that involve a major obstruction to air flow (stops and fricatives). We further will disregard the so-called laryngals, which have a major impact on the plausibility of the system, but are not essential to the emerging PGmc. system.

The obstruent development from PIE to PGmc. completely relies on redistributions of phonetic features with the sole motivation to build up a balanced structure. Empty structural slots attract sound clusters. Filling in these gaps, new slots emerge and a process comparable to a chain reaction is triggered. Thus in structural linguistic terminology such sequences of sound shifts are known as a drag chain.

Now you may ask, what exactly initiated the chain reaction. Well, structuralists are not much concerned with this kind of questions. Language systems are never perfect, so they are subject to continual change and it was just a matter of coincidence, that Gmc. languages arose. Of course, I myself sought an explanation in the pride of Germanic tribes, fostering their own indentity in language. Of course in the 19th century such thoughts were rather popular!

Anyway, at the end of my considerations I would like to come back to our topic - cannabis and hemp. According to the famous historian Herodotus (Histories IV) the Greek expression itself was loaned from the Scythian language in the 5th century BC. Now Gmc. people are not supposed to have known the plant and the word before this time. But consonant shift k > x(h) and b > p being attested, it can be claimed that around or after the 5th century BC Grimm's Law was still active.

In the 2nd century however, when first Latin loanwords seeped into Gmc. languages, Grimm's Law was already ineffective. ModE. tile < Lat. tegula bears no shifting of /t/. So our cannabis-hemp pair offers a substantial clue for the dating of the relative chronology of Grimm's Law.

With these final remarks in mind I would like to close the tutorial. I hope, you enjoyed my presentation. For your practice I have prepared a short exercise that you should be able to carry out by now. Please be so kind and hand over the results to your instructor when you are done.

With kind regards,

JG