In the beginning, all that was available to humans to observe the Cosmos was the human eye. However, the Universe is a much bigger place that the human eye can sense. For instance, on a very clear night, the human eye can only see 3,000 - 4,000 stars, even tho trillions exist.
In addition, as a detector, the human eye has a number of limitations:
this means you can't see faint stars.
and most of the interesting physics in the Universe
radiates at other wavelengths.
the eye doesn't integrate. Photons are destroyed
upon impact with your retina
information
transmission in the universe involves particle destruction.
your brain is an inefficient archive.
more on this later.
To do better than the eye, we need to build a light bucket (e.g. a big
mirror). The more area, the more light can be collected and therefore
fainter objects can be seen.
In simple terms, we construct a mirror that collects light and reflects
it to a point of focus. At the focus point, we could put a detector.

The kinds of detectors available to us to place at the focal point are:
These days we can make really large mirrors.

8 meter diameter mirror

and build a network of these telescopes
But ultimately, there are serious limitations
to building telescopes on the ground (or on mountaintops).
We can see fine detail (about 20 times finer) by avoiding imaging
astronomical sources through an atmosphere.
Image resolution is limited by motions
in our atmosphere. Therefore, detailed precise measurements can
not be made. This is why we have the Hubble Space Telescope:

Also, the night sky on the Earth is getting
brighter due to aerosals and pollutants. A brighter background
makes faint object detection difficult (as we will see in the
Java simulations to come).
Detector Issues:
A Comprehensive Reference