I have wondered that myself from time to time and my response is speculative. For one thing, many plants with thorns of that type are scramblers and climbers, and the thorns invariably are so placed that they aid the climbing process and stabilise the grip.However, true as that may be, it is by no means definitive. Many trees, including one-time members of the genus Acacia (currently in the process of being split into multiple genera) have just such thorns. Some of them are what you might call semi-scramblers and their thorns often are rather fine, and incredibly sharp. You go near such a plant at your peril. The same applies to certain species of asparagus in southern Africa and no doubt elsewhere.
As for the protective effect of the thorns, I can only imagine that it is more difficult for selective browsers to avoid such small, affectionate thorns, than to extract leaves from between large straight thorns. A tiny, delicate antelope that could easily retrieve individual leaves from between large acacia thorns, might have a far nastier experience with say Acacia melanacantha var detinens or Asparagus stipulaceus, than with the giant thorns of some strains of Acacia grandicornuta. The victim might well find the risks of feeding from plants with tiny hooked thorns too great to be worth while except in times of drought.
Black rhinoceros might take a different view. This could affect the distribution of various thorny species relative to the occurrence of various classes of browsers.
I've seen lambs so tangled up in brambles that they can hardly move. Perhaps it's actually a carnivorous plant that we haven't yet recognised as such!
It does seem odd, though. The main consumers of brambles are too small to care about the thorns (mites, insects, dormice, birds etc.). The one thing that many thorny plants seem to have in common is forming dense thickets, so perhaps the deterrent effect of hooks helps establish an untrodden area for growth. Bramble also readily roots from the stem, making loops of thorny growth that can be very hard to walk through.
I had actually speculated on the carnivorous idea, but I cannot think of a convincing example.
There are examples of plants and animals going to war, sometimes fatally in the long run, but very often they come to some kind of arrangement. For instance, rabbits can make a regular impregnable warren of a bramble tangle, but wthout killing it, whereas their droppings pay for the damage they do to the brambles.
"The main consumers of brambles are too small to care about the thorns (mites, insects, dormice, birds etc.)." - Maybe, the thorns are not there to protect the plant itself but the animals that feed on these consumers? Just speculating...
Roses and brambels are climbing plants. Instead of using a sticky glue type substance they hook into their environment with backward facing thorns which gives them the necessary grip to continue forward. The torns are not there for animals at all, they are not protective they are not offensive.
I should add that they have defensive side effects, such as preventing herbivores from eating the branches but next time you are out look at the new shoots, which herbivores like more then established shoots, and you will notice they do not have torns, the thorns take a while to develop. The PRIMARY reason for thorns on these plants is grip and their climbing ability.
There are some interesting answers but they skirt around the issue of a "carnivorous" plant. It may be that brambles, for instance, Himilaya blackberry bramble patches are indeed carnivorous. It would work for the plants if animals became stuck and die causing the animals to decay and add nutrients to the soil. Also, dead animals would attract other animals that would either become stuck themselves or find a home living in the brambles. That would add even more nutrients. There are many instances in nature of baiting. Once smaller animals, raccons for instance, make their safe home in bramble patches the bramble plants start enjoying the benefits of the excrement and other nutrient releasing material brought in by those smaller animals.
Birds do a great job of spreading seeds from berry producing brambles. The need is for nutrients as the "seeding" problem is solved by the birds and other berry eating animals that spread both seeds and nutrients via their excrement.
Roses were among the earliest flowering plants and possibly evolved thorns for defence against broad-snouted megafauna grazers; e.g., with rhinoceros-style noses. The more gracile giraffe/antelope style snouts came later but the thorns continue to be useful for defence for both the plants and the animals which they shelter. And because sheltered (and captured!) animals contribute nutrients its win-win.
Just adding this to the discussion, not as an answer - notice that thorns work on fur, not on feathers (or scales). Birds don't get tangled in them. So a bramble could have a very different evolutionary relationship with birds than with mammals.
I have some reservations on this. Your points are good, but I do not accept that thorns work on "fur" in general. Hooked thorns certainly work on wool, or any form of matted hairiness in which they can catch, but I have yet to see an otter, wildcat, or wild rabbit caught by its fur.
As for the different relationship between the likes of brambles and mammals or birds, this is largely true too, but not as clearly as the relationship between say chillis and birds in preference to mammals. Solanaceous plants tend to have soft seeds that are destroyed by the chewing of mammals, but pass unharmed through the guts of typical fruit-eating birds.
Brambles however have hard-coated seeds, many of which survive chewing. Very likely they are spread mainly by birds, but mammals such as pigs, deer, bear, rabbit, and humans also are effective vectors.
The range of cooperation or partial cooperation between various species of plants and various species of animals is wide and not easily to be categorised.
But, as I said, your points certainly have merit and deserve thought before we jump to any conclusions.