 Welcome back. Here is another morphological tree I want to generate with you using Liz Selkirk's proposals concerning a compositional structure of words. This has turned out to be very helpful for the analysis of present-day English derivatives. So here is another derivative for you. It is overgeneralization and this will be the result of our analysis. But how do we arrive at such a result? Well, let us again build the morphological tree step by step. As usual, you should start using an empty sheet with the word you have to analyze written down at the bottom so that you have enough space on top of it. Because we are generating a tree structure right here. Step one, as usual in such morphological analysis, exercises is you provide a morphological analysis. A simple analysis that is the analysis of the derivative into its base form and into its affixes. For reasons of readability, let us normalize all morphs orthographically or phonologically. That is, we use the defaults, if possible. For example, ION for, in this case, realization, generalization. So instead of the alomorphation, we are using the default morph ION. The next step is we assign word classes to all the morphs. General is an adjective. I's turns adjectives into verbs. So the inherent word class is verb and ION is nominal is a noun. Well, it is not really a noun, but it turns verbs into nouns. The problem is over. Over has, as we will see in a second, no real constant or permanent or inherent word class. Okay, let's now start generating the tree structure. We start with the base form and combine it with the suitable affix to the left and right. And of course, the affix we can combine, we can take for the combination with general is I's. Over general is not a word. You wouldn't find it in the corpora such as the British National Corpus. So we have to combine I's with general. And now we can ask the question, what is the result? What is the category that has to be associated with the mother node? Well, to do this, we have to take the percolation conventions that were defined in the 1980s by Selkirk. Now, here they are again. And in this case, we can simply apply rule number one, which passes the features of the affix up to the first branching node. So we have this situation generalize is clearly a verb. Now we take this new base form generalize and combine it with another suitable affix. The problem here is that both affixes are suitable. There's a verb such as over generalize. There is also a noun like generalization. Well, what can we do? Well, we could perhaps quickly look the resulting word forms up in the British National Corpus. And then we will find that over generalization is more frequent than over generalize, which, by the way, is spelled without a hyphen. So let's take generalization next. And this is then the new branching situation. And again, we need syntactic features for the new mother node. And we can simply pass the features of the affix up to the new branching node. So generalization is a noun. OK. And now we have to combine the prefix over with this new word form, with this new derivative generalization. And it is, of course, this structure. And again, we have to assign syntactic features to the new mother node. That is, OK, let's apply rule number one of the percolation conventions, which looks like this. The problem, however, is that the prefix over, as I've already said, has no real features. It can be attached to all sorts of base forms, such as overdue, overkill, overoptimistic, and so on and so forth. Hence, the mother node fails to obtain syntactic features. So we cannot really apply rule number one. It doesn't work. So what can we do now? Well, we take rule number two. If a branching node fails to obtain features, well, then we take the feature of the second daughter, which is in this case the noun. And we then have this new situation where rule number two of the percolation conventions now can be applied. And the final result is that overgeneralization is a noun. OK, that's it. The print version of the solution is available in the VLC e-lecture library. Well, and if you want to see more trees, you have to join us and take part in our practicals or look more solutions up in our practicals, in our practical sheets, where we have more than 50 trees with structures of this kind. Thanks and see you again.