Tonsil Cancers are quite sensitive to radiation and in the early stages have an excellent control rate with radiation alone. The literature review by Parsons  showed radiation had identical cure rates with surgery but much lower rates of complications (see Parsons.) Read review here.

Note that much of the data below was before the use of combined chemo-radiation and modern results are much higher (especially in stage III and IVa). Whereas most patients are now treated with combination of chemotherapy and high dose radiation using IMRT techniques (which has resulted in much better survival statistics, go here and here and here and here and here and here, esp if HPV + go here). So for advanced oropharynx cancer the survival rates have need pushed up form the 30 to 50% range now up to the 70 to 90% range,

Lymphoma of the tonsil is discussed here.

Consider the following information:
NCCN Guidelines: early stage, intermediate stage, advanced stage and recommended radiation dose.

Cure Rates with Radiation Alone
Cure Rates comparing radiation with surgery
more local control data, regional control data and survival data
Local Control Data and Survival Data
Sites of local spread, nodes at risk,
Typical small radiation fields: port1 ,  port 2 ,  port 3 , port 4, port 5, isodose
Typical fields for more advanced tonsil cance
picture of squamous cancer in tonsil, picture #2 and picture #3,
picture of lymphoma in tonsil / PET scan , more images here

There is some debate as to which is better, surgery or radiation for tonsil cancer, as the study notes below from the University of Florida, many of these patients can be treated entirely with just radiation. The literature review by Parsons also showed radiation had identical cure rates with surgery but much lower rates of complications (see Parsons.)

Radiation Therapy for Squamous Cell Carcinoma of the Tonsillar Region: A Preferred Alternative to Surgery?

By William M. Mendenhall, Robert J. Amdur, Scott P. Stringer, Douglas B. Villaret, Nicholas J. Cassisi

From the Departments of Radiation Oncology and Otolaryngology, University of Florida College of Medicine, Gainesville, FL. Jurnal of Clinical Oncology, Vol 18, Issue 11 (June), 2000: 2219-2225

PURPOSE: There are no definitive randomized studies that compare radiotherapy (RT) with surgery for tonsillar cancer. The purpose of this study was to evaluate the results of RT alone and RT combined with a planned neck dissection for carcinoma of the tonsillar area and to compare these data with the results of treatment with primary surgery.
PATIENTS AND METHODS: Four hundred patients were treated between October 1964 and December 1997 and observed for at least 2 years. One hundred forty-one patients underwent planned neck dissection, and 18 patients received induction (17 patients) or concomitant (one patient) chemotherapy. All patients were treated with curative intent with a continuous course of once-daily (160 patients) or twice-daily (240 patients) RT. Patients treated once daily usually received 1.80 to 2 Gy per fraction to a total dose of 47.15 Gy to 80 Gy (median, 65.65 Gy). Patients treated twice daily received 1.20 Gy per fraction to doses ranging from 66.90 Gy to 87.70 Gy (median, 76.80 Gy), with a 4- to 6-hour interfraction interval. A minimum 6-hour interfraction interval has been used in recent years. Forty-five patients treated with the planned split-course technique that was used between 1970 and 1974 were excluded.After external beam RT, one hundred seven patients underwent an interstitial brachytherapy boost to the primary tumor site, using radium needles, cesium needles, or iridium-192 hairpins. Twenty patients received part of their treatment with intraoral cone irradiation, using 250 kVp x-rays. Seventeen patients with advanced disease treated in recent years received two to three cycles of fluorouracil and cisplatin induction chemotherapy before RTone additional patient received concomitant chemotherapy. One hundred forty-one patients underwent planned neck dissection as part of their treatment.The neck dissection preceded RT in 16 patients and followed RT in 125 patients.

RESULTS: Five-year local control rates, by tumor stage, were as follows: T1, 83%; T2, 81%; T3, 74%; and T4, 60%. Multivariate analysis revealed that local control was significantly influenced by tumor stage (P = .0001), fractionation schedule (P = .0038), and external beam dose (P = .0227). Local control after RT for early-stage cancers was higher for tonsillar fossa/posterior pillar cancers than for those arising from the anterior tonsillar pillar. Five-year cause-specific survival rates, by disease stage, were as follows: I, 100%; II, 86%; III, 82%; IVa, 63%; and IVb, 22%. Multivariate analysis revealed that cause-specific survival was significantly influenced by overall stage (P = .0001), planned neck dissection (P = .0074), and histologic differentiation (P = .0307). The incidence of severe late complications after treatment was 5%.

CONCLUSION: RT alone or combined with a planned neck dissection provides cure rates that are as good as those after surgery and is associated with a lower rate of severe complications.