Well... it depends. "Color" can mean different things depending on who is using the word and how it is be used. Physics would tell us that the color of an object is defined by the various wavelengths of electromagnetic radiation emitted by the object. Chemistry would tell us that the color of an object can be further refined as the relative sum of wavelengths emitted by each of the different pigments within the object. Neurobiology would tell us that the color of an object is actually the sum of inputs from the cone cells in the retina, which doesn't always correlate directly with the physical wavelength model. Here's a list of previous responses on the subject that should be both helpful and confusing:

• How do atoms of an object absorb or reflect photons/ light to give color?
• Colors; wavelengths not adding up for me.
• What chemical properties cause things to be perceived a certain color?
• Can you formulate what color a molecule will reflect if only given its....
• How can purple exist?
• Why do we view red as the hottest color but hottest stars are blue?
• When colour fades where does it go?
• How can you measure the amount of pigment in plants?
• Does pure water has colour?

The final answer to your (and your friend's) question is that it depends on what each of you mean by "color". If one defines the color of an object by the nature of its pigments, or more precisely, by how its pigments act under white light - an apple is red if it contains "red" pigments - then the color of the object is absolute, regardless of the light source. On the other hand, if one defines the color of an object by its reflective spectrum under specific conditions - apples are black under blue lights because they do not reflect this light - then the only absolute color an object can have is black, since it is black under all possible lighting conditions. Linguistically, even scientists will use both definitions interchangeably. For example, many physics professors will talk about "shining a blue light on a red ball..." which would be an absolute usage - "red ball" refers to a ball that is red under white light - but then say that "under blue light the ball is black," which would be a conditional usage.

Confounding all of this is fluorescence: the ability of many pigments to emit specific wavelengths of visible light that are different from the wavelengths that they absorb. (Reflection is emission of the same wavelength of visible light as that absorbed.) The most common examples of fluorescent pigments are those that glow different colors under black lights; that is, they absorb UV irradiation and emit visible light. These can only be described conditionally, since their colors are not visible under white light - if one went into a store to purchase fluorescent paint and used the absolute description, white or colorless, one would come away with paint that probably didn't fluoresce. In biology, we commonly use fluorochromes, as we call fluorescent pigments, to label cells for microscopy, and then refer to the glowing cells as "green" or "red", even though these colors are specific to our microscopic conditions.

Ultimately, most scientists would agree that the conditional definition is more accurate, regarding the nature of light and the functions of pigments and cone cells, but that the absolute definition is often more useful, where the "under white light" condition is understood, because "under white light" is our most common frame of reference.