As far as reality is concerned, it has to begin with what can be seen and touched. In a nutshell, it should begin with sensorial data. In a physicalist account, these sensorial data have to be grounded in physical properties. It is a long tradition in the West that one envisages an ontic object that holds together all kinds of sensorial properties of this (maybe conceived) object, such as smell, taste, shape and color. For example, we think of the smell, shape, and color of a rose instead of disconnected sensorial data. In classical physics, these sensorial properties are taken as something that can be derived from more elementary properties. For example, temperature is the macroscopic manifestation of molecular velocity. Specifically, length, mass, time and charge consists of the elementary physical properties of a classical physical object (a rigid body). Other properties such as momentum, velocity, acceleration, energy, etc. can then be derived from the elementary properties. We notice that the Cartesian coordinates can also be envisaged as three mutually perpendicular rigid yardsticks of infinite length.
In a more modern form (mainly due to the Theory of Relativity), all physical properties have to rely on operational definitions. That is, for every physically meaningful property there must be a physically feasible operation that can be performed and will reveal this property. For example, simultaneity can be established only by synchronizing two clocks with a physically feasible medium -- light flashes, from which the property of time can be defined. Combining this with the world view above, an operation has to be performed on something, and this something can be regarded as an ontic ``physical object.'' Intuitively, physical reality consists of all the objects that manifest physical properties. The objectivity of physical reality guarantees that even if no observer is present, all physical properties are well defined and stable. Although not necessarily leading to materialism, a world view like this is classical realist. In short, in the classical view something is real if and only if it is operationally well defined. Furthermore, in the classical world view, the universe is well defined.
However, in quantum theory, this naive view has to be revised. In discussing quantum objects, all properties (represented by mathematical symbols) are physical in that they can indeed be ``seen'' and ``touched'' if one performs proper measurements. However, we have to abandon the naive assumption of certain ontic objects. Nevertheless, we can still hold these properties as ``real,'' if we extend our understanding of ``real'' a little bit. As to what a ``real'' stone means in everyday life -- when it is ``kicked,'' it ``kicks back,'' and the quantum objects do ``kick back'' and can be therefore taken as real. The ``kicking'' in quantum mechanics, however, is accomplished by performing measurement. But what ``kicks'' back in the quantum world depends on the arrangement of measuring devices (we assume quantum mechanics is correct in this regard). The merit of this view is that real does not have to be a synonym of naively objective.
If we stick to this idea of reality and apply it to language, linguistic objects can be taken as real because when they are ``kicked,'' they ``kick back,'' as in physical reality. However this time what kicks back seems to be the ``meaning'' of the symbol. Moreover, the ``meaning'' is embedded in the concept generated by an aggregate of symbols. Since quantum theory is our best theory of physics, we have to resort to quantum measurement to account for a physicalist process of ``kicking'' as well.
Let us see where this position leads to. As in quantum theory, meaning can be considered neither symbol nor concept but as a measuring process. The measurement, however, is active and participatory in that a human arranges his/her instrument in a particular orientation and chooses to perform a particular measurement at a specific place and time. Therefore, a symbol that is supposed to point to an apparently ``objective'' concept, such as ``smoke means fire,'' has nevertheless intentional property. It is the ``kicking-back'' that answers in response to measurement. In quantum mechanical terms, the objectivity can be approximately regarded as the pureness and/or average of a superposed quantum state. In this sense, the objectivity is a quantitative property rather than a qualitative one.
Now we are ready to formally give a quantum theoretical account of language.